No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Predicting Intent Using Activity Logs: How Goal Specificity and Temporal Range Affect User Behavior
Abstract
People have different intents in using online platforms. They may be trying to accomplish specific, short-term goals, or less well-defined, longer-term goals. While understanding user intent is fundamental to the design and personalization of online platforms, little is known about how intent varies across individuals, or how it relates to their behavior. Here, we develop a framework for understanding intent in terms of goal specificity and temporal range. Our methodology combines survey-based methodology with an observational analysis of user activity. Applying this framework to Pinterest, we surveyed nearly 6000 users to quantify their intent, and then studied their subsequent behavior on the web site. We find that goal specificity is bimodal - users tend to be either strongly goal-specific or goal-nonspecific. Goal-specific users search more and consume less content in greater detail than goal-nonspecific users: they spend more time using Pinterest, but are less likely to return in the near future. Users with short-term goals are also more focused and more likely to refer to past saved content than users with long-term goals, but less likely to save content for the future. Further, intent can vary by demographic, and with the topic of interest. Last, we show that user's intent and activity are intimately related by building a model that can predict a user's intent for using Pinterest after observing their activity for only two minutes. Altogether, this work shows how intent can be predicted from user behavior.
... ey categorized user intents into knowledge orientation, mental image, navigation, and transaction; these intents describe search activities in Flickr, a digital photo sharing platform instead of a search engine. A er that, a two-dimensional taxonomy was proposed in [7]. However, this work was based on another sharing platform (Pinterest). ...
... We see that "late fusion" and "early fusion" perform very similarly. It is worth pointing out that in absolute terms our performance gures are similar to other intent classi cation tasks considered in the literature, such as [7,39], even though we use a much sparser signal than [7] and, unlike [39], only use features that can be collected in real-world scenarios. ...
... We see that "late fusion" and "early fusion" perform very similarly. It is worth pointing out that in absolute terms our performance gures are similar to other intent classi cation tasks considered in the literature, such as [7,39], even though we use a much sparser signal than [7] and, unlike [39], only use features that can be collected in real-world scenarios. ...
What are the intents or goals behind human interactions with image search engines? Knowing why people search for images is of major concern to Web image search engines because user satisfaction may vary as intent varies. Previous analyses of image search behavior have mostly been query-based, focusing on what images people search for, rather than intent-based, that is, why people search for images. To date, there is no thorough investigation of how different image search intents affect users' search behavior. In this paper, we address the following questions: (1)Why do people search for images in text-based Web image search systems? (2)How does image search behavior change with user intent? (3)Can we predict user intent effectively from interactions during the early stages of a search session? To this end, we conduct both a lab-based user study and a commercial search log analysis. We show that user intents in image search can be grouped into three classes: Explore/Learn, Entertain, and Locate/Acquire. Our lab-based user study reveals different user behavior patterns under these three intents, such as first click time, query reformulation, dwell time and mouse movement on the result page. Based on user interaction features during the early stages of an image search session, that is, before mouse scroll, we develop an intent classifier that is able to achieve promising results for classifying intents into our three intent classes. Given that all features can be obtained online and unobtrusively, the predicted intents can provide guidance for choosing ranking methods immediately after scrolling.
... User engagement has previously been modeled based on repeat consumption patterns [11,13], binge watching [45], and patterns of switching between boredom and sensitization [22]. A user's intent in using a particular application can be predicted from the user's behavior [14,28] since it is strongly influenced by their intent [2]. Typical modeling of user engagement has considered users to have a single "lifetime" during which the user typically becomes less and less engaged on the platform [15,48]. ...
... Previous studies have shown that user engagement can be increased through notifications [3], incentives including badges [9,10], gamification [8,18,20], and social network features [5,20]. Furthermore, understanding the user's intent [14,28] can help designers surface different interaction modes [39], provide better contextual help [40], and personalize search results and recommendations [41,42]. This work. ...
Mobile health applications that track activities, such as exercise, sleep, and diet, are becoming widely used. While these activity tracking applications have the potential to improve our health, user engagement and retention are critical factors for their success. However, long-term user engagement patterns in real-world activity tracking applications are not yet well understood. Here we study user engagement patterns within a mobile physical activity tracking application consisting of 115 million logged activities taken by over a million users over 31 months. Specifically, we show that over 75% of users return and re-engage with the application after prolonged periods of inactivity, no matter the duration of the inactivity. We find a surprising result that the re-engagement usage patterns resemble those of the start of the initial engagement period, rather than being a simple continuation of the end of the initial engagement period. This evidence points to a conceptual model of multiple lives of user engagement, extending the prevalent single life view of user activity. We demonstrate that these multiple lives occur because the users have a variety of different primary intents or goals for using the app. We find evidence for users being more likely to stop using the app once they achieved their primary intent or goal (e.g., weight loss). However, these users might return once their original intent resurfaces (e.g., wanting to lose newly gained weight). Based on insights developed in this work, including a marker of improved primary intent performance, our prediction models achieve 71% ROC AUC. Overall, our research has implications for modeling user re-engagement in health activity tracking applications and has consequences for how notifications, recommendations as well as gamification can be used to increase engagement.
... As a result, the same recommendations will be offered when she returns next time. However, user preferences may change over time [7,26]. The time elapse effect on user preferences should be considered and thus the preference representation needs to be updated to the query time. ...
... The continuous nature of our model allows it to track the evolution of the underlying system from irregular observations, and no longer need the equal-length slice segmentation of the whole timeline, which empowers our method to perceive more finegrained temporal information compared with previous methods. We can then apply various ODE solvers to integrate the ODE function in Eq. (7). ODE solvers discretize time variable and convert an integral into many steps of additions. ...
Session-based recommendation (SBR) aims to predict the user next action based on the ongoing sessions. Recently, there has been an increasing interest in modeling the user preference evolution to capture the fine-grained user interests. While latent user preferences behind the sessions drift continuously over time, most existing approaches still model the temporal session data in discrete state spaces, which are incapable of capturing the fine-grained preference evolution and result in sub-optimal solutions. To this end, we propose Graph Nested GRU ordinary differential equation (ODE), namely GNG-ODE, a novel continuum model that extends the idea of neural ODEs to continuous-time temporal session graphs. The proposed model preserves the continuous nature of dynamic user preferences, encoding both temporal and structural patterns of item transitions into continuous-time dynamic embeddings. As the existing ODE solvers do not consider graph structure change and thus cannot be directly applied to the dynamic graph, we propose a time alignment technique, called t-Alignment, to align the updating time steps of the temporal session graphs within a batch. Empirical results on three benchmark datasets show that GNG-ODE significantly outperforms other baselines.
... The understanding and prediction of users' intents is an essential component for personalized search experience [12,44,48]. In other domains, such as arts and fashion [10], and psychology [15,38], user intentions were also investigated and were shown to be predictable from behavior logs [10]. In the context of recommendation systems, prior work leveraged interactive systems to elicit signals about user intentions, such as conversation-based [25], survey-based [52], and critiquebased [9] systems. ...
... The understanding and prediction of users' intents is an essential component for personalized search experience [12,44,48]. In other domains, such as arts and fashion [10], and psychology [15,38], user intentions were also investigated and were shown to be predictable from behavior logs [10]. In the context of recommendation systems, prior work leveraged interactive systems to elicit signals about user intentions, such as conversation-based [25], survey-based [52], and critiquebased [9] systems. ...
People's content choices are ideally driven by their intentions, aspirations, and plans. However, in reality, choices may be modulated by recommendation systems which are typically trained to promote popular items and to reinforce users' historical behavior. As a result, the utility and user experience of content consumption can be affected implicitly and undesirably. To study this problem, we conducted a 2 × 2 randomized controlled field experiment (105 urban college students) to compare the effects of intention informed recommendations with classical intention agnostic systems. The study was conducted in the context of spoken word web content (podcasts) which is often consumed through subscription sites or apps. We modified a commercial podcast app to include (1) a recommender that takes into account users' stated intentions at onboarding, and (2) a Collaborative Filtering (CF) recommender during daily use. Our study suggests that: (1) intention-aware recommendations can significantly raise users' interactions (subscriptions and listening) with channels and episodes related to intended topics by over 24%, even if such a recommender is only used during onboarding, and (2) the CF-based recommender doubles users' explorations on episodes from not-subscribed channels and improves satisfaction for users onboarded with the intention-aware recommender.
... User engagement has previously been modeled based on repeat consumption patterns [11,13], binge watching [45], and patterns of switching between boredom and sensitization [22]. A user's intent in using a particular application can be predicted from the user's behavior [14,28] since it is strongly influenced by their intent [2]. Typical modeling of user engagement has considered users to have a single "lifetime" during which the user typically becomes less and less engaged on the platform [15,48]. ...
... Previous studies have shown that user engagement can be increased through notifications [3], incentives including badges [9,10], gamification [8,18,20], and social network features [5,20]. Furthermore, understanding the user's intent [14,28] can help designers surface different interaction modes [39], provide better contextual help [40], and personalize search results and recommendations [41,42]. This work. ...
Mobile health applications that track activities, such as exercise, sleep, and diet, are becoming widely used. While these activity tracking applications have the potential to improve our health, user engagement and retention are critical factors for their success. However, long-term user engagement patterns in real-world activity tracking applications are not yet well understood. Here we study user engagement patterns within a mobile physical activity tracking application consisting of 115 million logged activities taken by over a million users over 31 months. Specifically, we show that over 75% of users return and re-engage with the application after prolonged periods of inactivity, no matter the duration of the inactivity. We find a surprising result that the re-engagement usage patterns resemble those of the start of the initial engagement period, rather than being a simple continuation of the end of the initial engagement period. This evidence points to a conceptual model of multiple lives of user engagement, extending the prevalent single life view of user activity. We demonstrate that these multiple lives occur because the users have a variety of different primary intents or goals for using the app. We find evidence for users being more likely to stop using the app once they achieved their primary intent or goal (e.g., weight loss). However, these users might return once their original intent resurfaces (e.g., wanting to lose newly gained weight). Based on insights developed in this work, including a marker of improved primary intent performance, our prediction models achieve 71% ROC AUC. Overall, our research has implications for modeling user re-engagement in health activity tracking applications and has consequences for how notifications, recommendations as well as gamification can be used to increase engagement.
... Modeling User Behavior. With the advent of e-commerce, much work has been devoted to understanding behavior of online users [1,8], and specifically predicting purchase behaviors [5,7]. As the former line of work, Lo et al. [8] studied user activity and purchasing behaviors that vary over time, especially focusing on user purchasing intent. ...
... As the former line of work, Lo et al. [8] studied user activity and purchasing behaviors that vary over time, especially focusing on user purchasing intent. Most recently, Cheng et al. [1] extended [8] by generalizing their analysis on characterizing the relationship between a user's intent and his behavior. Our goal is different in that rather than predicting users' various intents from their online behaviors, we focus on predicting users' future purchase. ...
Understanding users' behavior and predicting their future purchase are critical for e-commerce companies to boost their revenue. While explicit user feedback such as ratings plays the most significant role in eliciting users' preferences, such feedback is scarce, which prompts the need for leveraging more abundant implicit user feedback such as purchase record. Consequently, recent studies focused on leveraging users' past purchase record to predict their purchase. However, their performance is not satisfactory due to 1) the lack of purchase history of users, and 2) more importantly the ill-posed assumption of non-purchased items equally being considered as negative feedback. In this paper, we define new pairwise relationships among items aiming at overcoming the limitations of existing works, and propose a novel method called P3S that stands for modeling pairwise relationships among three disjoint item sets, which leverages users' click record in conjunction with their purchase record. Precisely, we partition the items into three disjoint sets based on users' purchase and click record, define new pairwise relationships among them with respect to users, and reflect these relationships into our pairwise learning-to-rank method. Experiments on two real-world datasets demonstrate that our proposed method outperforms the state-of-the-art baselines in the task of predicting users' future purchase in e-commerce.
... Research has been focused on creating models robust to noise in session data (Agichtein et al., 2006), and using a recurrent neural network to predict customer behaviour (Lang and Rettenmeier, 2017). Sismeiro and Bucklin (2004) predict purchasing task completion for a given user who completed at least one task earlier, whereas Cheng et al. (2017) explore user behavior on a content discovery platform to determine intent specificity and time in the future when a purchase is estimated to take place. Some work in the field focuses on using queries for purchasing behavior modeling. ...
In this thesis, we investigate multimodal machine learning for information retrieval, focusing on vision and language. Our research is organized into three main areas: (i) dense and sparse cross-modal retrieval, where we examine reproducibility issues and propose methods for learned sparse retrieval; (ii) representation learning and evaluation, where we study the limitations of vision-language contrastive learning and advocate for more robust evaluation frameworks; and (iii) product retrieval, where we analyze user behavior and leverage multimodal data to improve retrieval performance across categories of varying granularity.
... Ideally, such investigations should be supported either by an underlying theory, e.g., from psychology as in [36,135,139], or by empirical studies or domain-specific analyses, as done, e.g., for the music and video domains in [11,90,109,110,131]. A better understanding of these application-specific intents then serves as a basis for improved models for predicting the user's intent from their observed interactions, see, e.g., [22] using observational and survey data at Pinterest for improved intent prediction. ...
Many modern online services feature personalized recommendations. A central challenge when providing such recommendations is that the reason why an individual user accesses the service may change from visit to visit or even during an ongoing usage session. To be effective, a recommender system should therefore aim to take the users’ probable intent of using the service at a certain point in time into account. In recent years, researchers have thus started to address this challenge by incorporating intent-awareness into recommender systems. Correspondingly, a number of technical approaches were put forward, including diversification techniques, intent prediction models or latent intent modeling approaches. In this paper, we survey and categorize existing approaches to building the next generation of Intent-Aware Recommender Systems (IARS). Based on an analysis of current evaluation practices, we outline open gaps and possible future directions in this area, which in particular include the consideration of additional interaction signals and contextual information to further improve the effectiveness of such systems.
... Ideally, such investigations should be supported either by an underlying theory, e.g., from psychology as in [33,129,133], or by empirical studies or domain-specific analyses, as done, e.g., for the music and video domains in [9,82,102,103,124]. A better understanding of these application-specific intents then serves as a basis for improved models for predicting the user's intent from their observed interactions, see, e.g., [21] using observational and survey data at Pinterest for improved intent prediction. ...
Many modern online services feature personalized recommendations. A central challenge when providing such recommendations is that the reason why an individual user accesses the service may change from visit to visit or even during an ongoing usage session. To be effective, a recommender system should therefore aim to take the users' probable intent of using the service at a certain point in time into account. In recent years, researchers have thus started to address this challenge by incorporating intent-awareness into recommender systems. Correspondingly, a number of technical approaches were put forward, including diversification techniques, intent prediction models or latent intent modeling approaches. In this paper, we survey and categorize existing approaches to building the next generation of Intent-Aware Recommender Systems (IARS). Based on an analysis of current evaluation practices, we outline open gaps and possible future directions in this area, which in particular include the consideration of additional interaction signals and contextual information to further improve the effectiveness of such systems.
... Secondly, using median impressions viewed across all sessions as CustomerN has its limitations because it cannot account for additional variability within the same customer's sessions as seen in Figure 1. So we would like to segment customer sessions based on their mindset per session using same-session variables, historical activity, demographics, and geographical variables as detailed in [17,18] and subsequently calculate CustomerN as median impressions viewed at the Customer-Segment level. Lastly, we will monitor long-term KPIs to validate if improved Metric@CustomerN correlates with customer satisfaction and lifetime value. ...
Accuracy measures such as Recall, Precision, and Hit Rate have been a standard way of evaluating Recommendation Systems. The assumption is to use a fixed Top-N to represent them. We propose that median impressions viewed from historical sessions per diner be used as a personalized value for N. We present preliminary exploratory results and list future steps to improve upon and evaluate the efficacy of these personalized metrics.
... The user might have to look for flights, book a hotel and look for activities in the target location, which all represent potentially dissociated subtasks. It gets even more complicated when considering that different stages of information-need fulfilment might also have an impact on how a user behaves [13,19,46,214,222,227,265,275,279,301]. ...
Die Identifikation von Sessions zum Verständnis des Benutzerverhaltens ist ein Forschungsgebiet des Web Usage Mining. Definitionen und Konzepte werden seit über 20 Jahren diskutiert. Die Forschung zeigt, dass Session-Identifizierung kein willkürlicher Prozess sein sollte. Es gibt eine fragwürdige Tendenz zu vereinfachten mechanischen Sessions anstelle logischer Segmentierungen. Ziel der Dissertation ist es zu beweisen, wie unterschiedliche Session-Ansätze zu abweichenden Ergebnissen und Interpretationen führen. Die übergreifende Forschungsfrage lautet: Werden sich verschiedene Ansätze zur Session-Identifizierung auf Analyseergebnisse und Machine-Learning-Probleme auswirken? Ein methodischer Rahmen für die Durchführung, den Vergleich und die Evaluation von Sessions wird gegeben. Die Dissertation implementiert 135 Session-Ansätze in einem Jahr (2018) Daten einer deutschen Preisvergleichs-E-Commerce-Plattform. Die Umsetzung umfasst mechanische Konzepte, logische Konstrukte und die Kombination mehrerer Mechaniken. Es wird gezeigt, wie logische Sessions durch Embedding-Algorithmen aus Benutzersequenzen konstruiert werden: mit einem neuartigen Ansatz zur Identifizierung logischer Sessions, bei dem die thematische Nähe von Interaktionen anstelle von Suchanfragen allein verwendet wird. Alle Ansätze werden verglichen und quantitativ beschrieben sowie in drei Machine-Learning-Problemen (wie Recommendation) angewendet. Der Hauptbeitrag dieser Dissertation besteht darin, einen umfassenden Vergleich von Session-Identifikationsalgorithmen bereitzustellen. Die Arbeit bietet eine Methodik zum Implementieren, Analysieren und Evaluieren einer Auswahl von Mechaniken, die es ermöglichen, das Benutzerverhalten und die Auswirkungen von Session-Modellierung besser zu verstehen. Die Ergebnisse zeigen, dass unterschiedlich strukturierte Eingabedaten die Ergebnisse von Algorithmen oder Analysen drastisch verändern können.
... Inhibiting extensive app usage in those situations might better optimise user value. Another potential approach is proposed by Cheng et al. [49]. They construct a model that from two minutes of behavioural data can predict users' intention of the session, perhaps this model could be used to prevent erosion of intention during a usage session. ...
The tech industry has been criticised for designing applications that undermine individuals' autonomy. Recommender systems, in particular, have been identified as a suspected culprit that might exercise unwanted control over peoples' lives. In this article we try to assess the objectives of recommender system research and offer a nuanced discussion of how these objectives can align with users' goals. This discussion employs a qualitative literature survey connecting the dots between relevant research within the fields of psychology, design ethics, interaction design and recommender systems. Finally, we focus on the specific use-case of YouTube's recommender system and propose design changes that will better align with individuals' autonomy. Based on our analysis we offer directions for future research that will help secure rights to digital autonomy in the attention economy.
... Some previous works exploit LDA [2,12,13], n-gram [2,14] and deep learning model [15,16] to mine user intent from user behavior. At the same time, modeling user intents can help us understand user needs better and is significant in many scenarios; for instance, web searching [17,18], e-commerce application [4], image sharing social platform [19], activity tracking application [5] and recommander systems [20][21][22][23][24]. However, to our best knowledge, no related work has explicitly modeled user intent and its variation for differentiated user engagement forecasting in online social platforms. ...
User engagement prediction plays a critical role for designing interaction strategies to grow user engagement and increase revenue in online social platforms. Through the in-depth analysis of the real-world data from the world's largest professional social platforms, i.e., LinkedIn, we find that users expose diverse engagement patterns, and a major reason for the differences in user engagement patterns is that users have different intents. That is, people have different intents when using LinkedIn, e.g., applying for jobs, building connections, or checking notifications, which shows quite different engagement patterns. Meanwhile, user intents and the corresponding engagement patterns may change over time. Although such pattern differences and dynamics are essential for user engagement prediction, differentiating user engagement patterns based on user dynamic intents for better user engagement forecasting has not received enough attention in previous works. In this paper, we proposed a Dynamic Intent Guided Meta Network (DIGMN), which can explicitly model user intent varying with time and perform differentiated user engagement forecasting. Specifically, we derive some interpretable basic user intents as prior knowledge from data mining and introduce prior intents in explicitly modeling dynamic user intent. Furthermore, based on the dynamic user intent representations, we propose a meta predictor to perform differentiated user engagement forecasting. Through a comprehensive evaluation on LinkedIn anonymous user data, our method outperforms state-of-the-art baselines significantly, i.e., 2.96% and 3.48% absolute error reduction, on coarse-grained and fine-grained user engagement prediction tasks, respectively, demonstrating the effectiveness of our method.
... A user browsing session typically consists of a series of web pages accessed within a given time [59,60]. These sessions can be categorized into two main types: goal-oriented and general browsing [61,62]. To replicate a user's goal-oriented and general browsing, we create a list of legitimate URLs targeting popular news and streaming websites, including some random search engine queries. ...
We present an improved network attack evasion technique that allows malicious two-way communication and bypasses popular host and network intrusion techniques/systems that use deep packet inspection, signature analysis, and traffic behavior. The attack is based on previous research that leverages legitimate network traffic (existing or intuitively generated) from different contexts and reuses it to communicate malicious content. Still, contrary to previous research, the proposed approach: (i) provides increased bandwidth and allows us to exfiltrate large amounts of data with improved execution times while avoiding detection, and (ii) removes the administration privilege constraint that existed in previous implementations. Both novelties now make the attack feasible in real-world scenarios. We present two different attack implementations in different contexts, i.e., scripts/commands two-way communication and large data transfer. We test and validate our two implemented attacks using four popular NIDS, eight of the most popular endpoint protection solutions, and a Data Leakage Prevention System (DLP). Finally, we include a comparison of findings between our implementations of attacks and previous studies.
... There is a huge amount of information on the internet that contains various user intentions, such as search queries (Dai et al. 2006, Ashkan andClarke 2009), online forum posts , microblog posts (Wang et al. 2015), and activity logs (Cheng et al. 2017), which provide a potential source for detecting online user adoption intentions, such as the adoption of a product, service, or opinion (Fang et al. 2013). Among all types of adoption intentions, product adoption intentions contain significant business value for product providers. ...
Detecting product adoption intentions on social media could yield significant value in a wide range of applications, such as personalized recommendations and targeted marketing. In the literature, no study has explored the detection of product adoption intentions on social media, and only a few relevant studies have focused on purchase intention detection for products in one or several categories. Focusing on a product category rather than a specific product is too coarse-grained for precise advertising. Additionally, existing studies primarily focus on using one type of text representation in target social media posts, ignoring the major yet unexplored potential of fusing different text representations. In this paper, we first formulate the problem of product adoption intention mining and demonstrate the necessity of studying this problem and its practical value. To detect a product adoption intention for an individual product, we propose a novel and general multiview deep learning model that simultaneously taps into the capability of multiview learning in leveraging different representations and deep learning in learning latent data representations using a flexible nonlinear transformation. Specifically, the proposed model leverages three different text representations from a multiview perspective and takes advantage of local and long-term word relations by integrating convolutional neural network (CNN) and long short-term memory (LSTM) modules. Extensive experiments on three Twitter datasets demonstrate the effectiveness of the proposed multiview deep learning model compared with the existing benchmark methods. This study also significantly contributes research insights to the literature about intention mining and provides business value to relevant stakeholders such as product providers.
... Sismeiro and Bucklin [28] predict purchasing task completion for a given user who completed at least one task earlier, whereas Cheng et al. [7] explore user behavior on a content discovery platform to determine intent specificity and time in the future when a purchase is estimated to take place. Some work in the field focuses on using queries for purchasing behavior modeling. ...
The popularity of e-commerce platforms continues to grow. Being able to understand, model, and predict customer behaviour is essential for customizing the user experience through person-alized result presentations, recommendations, and special offers. Previous work has considered a broad range of prediction models as well as features inferred from clickstream data to record session characteristics, and features inferred from user data to record customer characteristics. So far, most previous work in the area of purchase prediction has focused on known customers, largely ignoring anonymous sessions, i.e., sessions initiated by a non-logged-in or unrecognized customer. However, in the de-identified data from a large European e-commerce platform available to us, more than 50% of the sessions start as anonymous sessions. In this paper, we focus on purchase prediction for both anonymous and identified sessions on an e-commerce platform. We start with a descriptive analysis of purchase vs. non-purchase sessions. This analysis informs the definition of a feature-based model for purchase prediction for anonymous sessions and identified sessions; our models consider a range of session-based features for anonymous sessions, such as the channel type, the number of visited pages, and the device type. For identified user sessions, our analysis points to customer history data as a valuable discriminator between purchase and non-purchase sessions. Based on our analysis, we build two types of predictors: (1) a predictor for anonymous sessions that can accurately predict purchase intent in anonymous sessions, beating a production-ready predictor by over 17.54% 1 ; and (2) a predictor for identified customers that uses session data as well as customer history and achieves an 1 of 96.20% on held-out data collected from a real-world retail platform. Finally, we discuss the broader practical implications of our findings.
... A second approach is to change the user interface based on a personalized prediction model. Recent work has found that classifers can be trained to predict these types of media use with high confdence, e.g., for Pinterest [24] and smartphone use [50]. For example, if YouTube expects that the user is visiting for ritualistic use, it could remain as is, or even go further to take control as in its Leanback mode for "efortless viewing" that autoplays a neverending stream of high-defnition recommendations [41]. ...
... A second approach then is an interface that is personalized for the user based on a prediction model. Recent work has found that classifiers can be trained to predict these types of media use with high confidence, e.g., for Pinterest [25] and smartphone use [51]. For example, if YouTube expects that the user is visiting for ritualistic use, it could remain as is, or even go further to take control as in its Leanback mode for "effortless viewing" that autoplays a neverending stream of high-definition recommendations [42]. ...
In the attention economy, video apps employ design mechanisms like autoplay that exploit psychological vulnerabilities to maximize watch time. Consequently, many people feel a lack of agency over their app use, which is linked to negative life effects such as loss of sleep. Prior design research has innovated external mechanisms that police multiple apps, such as lockout timers. In this work, we shift the focus to how the internal mechanisms of an app can support user agency, taking the popular YouTube mobile app as a test case. From a survey of 120 U.S. users, we find that autoplay and recommendations primarily undermine sense of agency, while search and playlists support it. From 13 co-design sessions, we find that when users have a specific intention for how they want to use YouTube they prefer interfaces that support greater agency. We discuss implications for how designers can help users reclaim a sense of agency over their media use.
... Click records with simple user-item interaction have been exploited for various tasks such as click-through rate (CTR) prediction in online advertising [22] and user intent prediction [23]. On purchase-relevant datasets, analysis over click records has been the most fundamental component of purchase prediction model. ...
Targeting at boosting business revenue, purchase prediction based on user behavior is crucial to e-commerce. However, it is not a well-explored topic due to a lack of relevant datasets. Specifically , no public dataset provides both price and discount information varying on time, which play an essential role in the user's decision making. Besides, existing learn-to-rank methods cannot explicitly predict the purchase possibility for a specific user-item pair. In this paper, we propose a two-step graph-based model, where the graph model is applied in the first step to learn representations of both users and items over click-through data, and the second step is a classifier incorporating the price information of each transaction record. To evaluate the model performance , we propose a transaction-based framework focusing on the purchased items and their context clicks, which contain items that a user is interested in but fails to choose after comparison. Our experiments show that exploiting the price and discount information can significantly enhance prediction accuracy.
... Sismeiro and Bucklin [28] predict purchasing task completion for a given user who completed at least one task earlier, whereas Cheng et al. [7] explore user behavior on a content discovery platform to determine intent specificity and time in the future when a purchase is estimated to take place. Some work in the field focuses on using queries for purchasing behavior modeling. ...
The popularity of e-commerce platforms continues to grow. Being able to understand, and predict customer behavior is essential for customizing the user experience through personalized result presentations, recommendations, and special offers. Previous work has considered a broad range of prediction models as well as features inferred from clickstream data to record session characteristics, and features inferred from user data to record customer characteristics. So far, most previous work in the area of purchase prediction has focused on known customers, largely ignoring anonymous sessions, i.e., sessions initiated by a non-logged-in or unrecognized customer. However, in the de-identified data from a large European e-commerce platform available to us, more than 50% of the sessions start as anonymous sessions. In this paper, we focus on purchase prediction for both anonymous and identified sessions on an e-commerce platform. We start with a descriptive analysis of purchase vs. non-purchase sessions. This analysis informs the definition of a feature-based model for purchase prediction for anonymous sessions and identified sessions; our models consider a range of session-based features for anonymous sessions, such as the channel type, the number of visited pages, and the device type. For identified user sessions, our analysis points to customer history data as a valuable discriminator between purchase and non-purchase sessions. Based on our analysis, we build two types of predictors: (1) a predictor for anonymous that beats a production-ready predictor by over 17.54% F1; and (2) a predictor for identified customers that uses session data as well as customer history and achieves an F1 of 96.20%. Finally, we discuss the broader practical implications of our findings.
... Thus, knowing the motive of the subject, they use it to predict actions for Prisoners Dilemma. The work in (Cheng, Lo, and Leskovec 2017) linked intent with goal specificity and temporal range when predicting intents in online platforms. Very recent research work uses deep-learning models for intent prediction (Qu et al. 2019;Pîrvu et al. 2018). ...
Artificial intelligence (AI) agents will need to interact with both other AI agents and humans. Creating models of associates help to predict the modeled agents' actions, plans, and intentions. This work introduces algorithms that predict actions , plans and intentions in repeated play games, with providing an exploration of algorithms. We form a generative Bayesian approach to model S#. S# is designed as a robust algorithm that learns to cooperate with its associate in 2 by 2 matrix games. The actions, plans and intentions associated with each S# expert are identified from the literature, grouping the S# experts accordingly, and thus predicting actions, plans, and intentions based on their state probabilities. Two prediction methods are explored for Prisoners Dilemma: the Maximum A Posteriori (MAP) and an Aggregation approach. MAP (≈ 89% accuracy) performed the best for action prediction. Both methods predicted plans of S# with ≈ 88% accuracy. Paired T-test shows that MAP performs significantly better than Aggregation for predicting S#'s actions without cheap talk. Intention is explored based on the goals of the S# experts; results show that goals are predicted precisely when modeling S#. The obtained results show that the proposed Bayesian approach is well suited for modeling agents in two-player repeated games.
... Thus, knowing the motive of the subject, they use it to predict actions for Prisoners Dilemma. The work in (Cheng, Lo, and Leskovec 2017) linked intent with goal specificity and temporal range when predicting intents in online platforms. Very recent research work uses deep-learning models for intent prediction (Qu et al. 2019;Pîrvu et al. 2018). ...
Artificial intelligence (AI) agents will need to interact with both other AI agents and humans. Creating models of associates help to predict the modeled agents' actions, plans, and intentions. This work introduces algorithms that predict actions, plans and intentions in repeated play games, with providing an exploration of algorithms. We form a generative Bayesian approach to model S#. S# is designed as a robust algorithm that learns to cooperate with its associate in 2 by 2 matrix games. The actions, plans and intentions associated with each S# expert are identified from the literature, grouping the S# experts accordingly, and thus predicting actions, plans, and intentions based on their state probabilities. Two prediction methods are explored for Prisoners Dilemma: the Maximum A Posteriori (MAP) and an Aggregation approach. MAP (~89% accuracy) performed the best for action prediction. Both methods predicted plans of S# with ~88% accuracy. Paired T-test shows that MAP performs significantly better than Aggregation for predicting S#'s actions without cheap talk. Intention is explored based on the goals of the S# experts; results show that goals are predicted precisely when modeling S#. The obtained results show that the proposed Bayesian approach is well suited for modeling agents in two-player repeated games.
... A different social media platform, Pinterest, is used to understand user intents in terms of temporal range and goal specificity [8]. The authors develop a framework which combines survey-based methodology with observational analysis of user activity. ...
... Commercial analysts and consultants have long studied user behavior data collection. Log (Cheng et al., 2017), search (Kim et al., 2015), payment (Liébana-Cabanillas et al., 2018), and click (Wang et al., 2017) behaviors of web or mobile users are the most common behavior studied among others. Recently, with the rise of emerging pervasive wearable and smart devices such as smartwatches and smartphones, the behavior data of movement (Vuković et al., 2018), eating , and sleeping (Alfeo et al., 2018) becomes possible to be collected. ...
In the face of current predicaments of facility management (FM), the concept of cognitive FM is proposed with a view to providing active intelligent management of a facility. In order to achieve such cognitive FM, how to integrate user behavior data into a cognitive FM system has to be solved. This paper serves as method guidance for it by putting forward the idea that location can serve as a gateway for the integration. Ultra-wideband (UWB) is recommended as the device layer to construct the 3D local positioning system for the cognitive FM system after comparison between different local positioning technologies from the accuracy, scalability, and cost dimensions. The way to bridge the user behavior data with facilities through coordinate transformation and location/distance computation is briefly introduced. Such of a uniform 3D coordinate system with high accuracy and scalability for FM situation can provide a common language for communication and computational applications. Finally, application scenarios for various facilities such as commercial building, office building, hospitals, warehouses, airports, and transportation stations are discussed.
... As a result, the same predictions and recommendations will be made to her regardless of when she returns. However, a user's intent changes over time [10] and thus her embedding needs to be updated (projected) to the query time. ...
Modeling sequential interactions between users and items/products is crucial in domains such as e-commerce, social networking, and education. Representation learning presents an attractive opportunity to model the dynamic evolution of users and items, where each user/item can be embedded in a Euclidean space and its evolution can be modeled by an embedding trajectory in this space. However, existing dynamic embedding methods generate embeddings only when users take actions and do not explicitly model the future trajectory of the user/item in the embedding space. Here we propose JODIE, a coupled recurrent neural network model that learns the embedding trajectories of users and items. JODIE employs two recurrent neural networks to update the embedding of a user and an item at every interaction. Crucially, JODIE also models the future embedding trajectory of a user/item. To this end, it introduces a novel projection operator that learns to estimate the embedding of the user at any time in the future. These estimated embeddings are then used to predict future user-item interactions. To make the method scalable, we develop a t-Batch algorithm that creates time-consistent batches and leads to 9x faster training. We conduct six experiments to validate JODIE on two prediction tasks---future interaction prediction and state change prediction---using four real-world datasets. We show that JODIE outperforms six state-of-the-art algorithms in these tasks by at least 20% in predicting future interactions and 12% in state change prediction.
... Another purpose of user behavior modeling leads to the discovery and prediction of user intention. Studies were done on predicting user intent and subsequent behavior, as well as prediction of a user's intention [8] and the users purchasing intent [17] on pinterest. Interesting to mention, some studies in the domain focus and models the consumption of a user [5,19]. ...
While mobile social apps have become increasingly important in people's daily life, we have limited understanding on what motivates users to engage with these apps. In this paper, we answer the question whether users' in-app activity patterns help inform their future app engagement (e.g., active days in a future time window)? Previous studies on predicting user app engagement mainly focus on various macroscopic features (e.g., time-series of activity frequency), while ignoring fine-grained inter-dependencies between different in-app actions at the microscopic level. Here we propose to formalize individual user's in-app action transition patterns as a temporally evolving action graph, and analyze its characteristics in terms of informing future user engagement. Our analysis suggested that action graphs are able to characterize user behavior patterns and inform future engagement. We derive a number of high-order graph features to capture in-app usage patterns and construct interpretable models for predicting trends of engagement changes and active rates. To further enhance predictive power, we design an end-to-end, multi-channel neural model to encode both temporal action graphs, activity sequences, and other macroscopic features. Experiments on predicting user engagement for 150k Snapchat new users over a 28-day period demonstrate the effectiveness of the proposed prediction models. The analysis and prediction framework is also deployed at Snapchat to deliver real world business insights. Our proposed framework is also general and can be applied to any online platform.
... As a result, the same predictions and recommendations will be made to her regardless of when she returns. However, a user's intent changes over time [10] and thus her embedding needs to be updated (projected) to the query time. The challenge here is how to accurately predict the embedding trajectories of users/items as time progresses. ...
Modeling sequential interactions between users and items/products is crucial in domains such as e-commerce, social networking, and education. Representation learning presents an attractive opportunity to model the dynamic evolution of users and items, where each user/item can be embedded in a Euclidean space and its evolution can be modeled by an embedding trajectory in this space. However, existing dynamic embedding methods generate embeddings only when users take actions and do not explicitly model the future trajectory of the user/item in the embedding space. Here we propose JODIE, a coupled recurrent neural network model that learns the embedding trajectories of users and items. JODIE employs two recurrent neural networks to update the embedding of a user and an item at every interaction. Crucially, JODIE also models the future embedding trajectory of a user/item. To this end, it introduces a novel projection operator that learns to estimate the embedding of the user at any time in the future. These estimated embeddings are then used to predict future user-item interactions. To make the method scalable, we develop a t-Batch algorithm that creates time-consistent batches and leads to 9x faster training. We conduct six experiments to validate JODIE on two prediction tasks---future interaction prediction and state change prediction---using four real-world datasets. We show that JODIE outperforms six state-of-the-art algorithms in these tasks by at least 20% in predicting future interactions and 12% in state change prediction.
... Another purpose of user behavior modeling leads to the discovery and prediction of user intention. Studies were done on predicting user intent and subsequent behavior, as well as prediction of a user's intention [8] and the users purchasing intent [17] on pinterest. Interesting to mention, some studies in the domain focus and models the consumption of a user [5,19]. ...
While mobile social apps have become increasingly important in people's daily life, we have limited understanding on what motivates users to engage with these apps. In this paper, we answer the question whether users' in-app activity patterns help inform their future app engagement (e.g., active days in a future time window)? Previous studies on predicting user app engagement mainly focus on various macroscopic features (e.g., time-series of activity frequency), while ignoring fine-grained inter-dependencies between different in-app actions at the microscopic level. Here we propose to formalize individual user's in-app action transition patterns as a temporally evolving action graph, and analyze its characteristics in terms of informing future user engagement. Our analysis suggested that action graphs are able to characterize user behavior patterns and inform future engagement. We derive a number of high-order graph features to capture in-app usage patterns and construct interpretable models for predicting trends of engagement changes and active rates. To further enhance predictive power, we design an end-to-end, multi-channel neural model to encode temporal action graphs, activity sequences, and other macroscopic features. Experiments on predicting user engagement for 150k Snapchat new users over a 28-day period demonstrate the effectiveness of the proposed models. The prediction framework is deployed at Snapchat to deliver real world business insights. Our proposed framework is also general and can be applied to other social app platforms.
... Wu et al. [49] used implicit feedback to identify two stages of searching: comparing search result pages and deciding to purchase on description pages. Researchers from Pinterest have revealed that goal specificity (e.g., visiting a web site for a specific goal) and temporal range (e.g., when the user envisions the goal will be achieved) influence user intent [10]. Su et al. [47] determined three types of product search intents in a commercial product search engine that also range in specificity and focus. ...
Music listening is a commonplace activity that has transformed as users engage with online streaming platforms. When presented with anytime, anywhere access to a vast catalog of music, users face challenges in searching for what they want to hear. We propose that users who engage in domain-specific search (e.g., music search) have different information-seeking needs than in general search. Using a mixed-method approach that combines a large-scale user survey with behavior data analyses, we describe the construct of search mindset on a leading online streaming music platform and then investigate two types of search mindsets: focused, where a user is looking for one thing in particular, and non-focused, where a user is open to different results. Our results reveal that searches in the music domain are more likely to be focused than non-focused. In addition, users' behavior (e.g., clicks, streams, querying, etc.) on a music search system is influenced by their search mindset. Finally, we propose design implications for music search systems to best support their users.
... Various cloud providers provision heterogeneous hosts that may significantly affect the performance of web-applications, which has a strong impact on user satisfaction [3]. Thus, we need to benchmark the hosts before deploying the real applications [6]. ...
Containerized web-applications have gained popularity recently due to the advantages provided by the containers including light-weight, packaged, fast start up and shut down and easy scalability. As there are more than 267 cloud providers, finding a flexible deployment option for containerized web-applications is very difficult as each cloud offers numerous deployment infrastructure. Benchmarking is one of the eminent options to evaluate the provisioned resources before product-level deployment. However, benchmarking the massive infrastructure resources provisioned by various cloud providers is a time consuming, tedious and costly process and is not practical to accomplish manually.
In this demonstration, we present Smart Docker Benchmarking Orchestrator (SmartDBO), a general orchestration framework that automatically benchmarks (deploys and executes) users' containerized web-applications across different cloud providers while meeting the constraints of budget and deployment configurations. SmartDBO aims to answer two questions: (i) how to automate the benchmarking of containerized web-application across multi-cloud environments?, (ii) how to maximize the diversity in a benchmarking solution which covers maximum numbers of cloud providers and types of provisioned infrastructures without exceeding users' budgets? We evaluate all the features of SmartDBO using SimplCommerce and TPC-W executing across Amazon AWS and Microsoft Azure.
Logged behavioral data is a common resource for enhancing the user experience on streaming platforms. In music streaming, Mehrotra et al. have shown how complementing behavioral data with user intent can help predict and explain user satisfaction. Do their findings extend to video streaming? Compared to music streaming, video streaming platforms provide relatively shallow catalogs. Finding the right content demands more active and conscious commitment from users than in the music streaming setting. Video streaming platforms, in particular, could thus benefit from a better understanding of user intents and satisfaction level. We replicate Mehrotra et al.’s study from music to video streaming and extend their modeling framework on two fronts: (i) improved modeling accuracy (random forests), and (ii) interpretability (Bayesian models).
Like the original study, we find that user intent affects behavior and satisfaction itself, even if to a lesser degree, based on data analysis and modeling. By proposing a grouping of intents into decisive and explorative categories we highlight a tension: decisive video streamers are not as keen to interact with the user interface as exploration-seeking ones. Meanwhile, music streamers explore by listening. In this study, we find that in video streaming, unsatisfied users provide the main signal: intent influences satisfaction levels together with behavioral data, depending on our decisive vs. explorative grouping.
YouTube has many features, such as homepage recommendations, that encourage users to explore its vast library of videos. However, when users visit YouTube with a specific intention, e.g., learning how to program in Python, these features to encourage exploration are often distracting. Prior work has innovated ‘commitment interfaces’ that restrict social media but finds that they often indiscriminately block needed content. In this paper, we describe the design, development, and evaluation of an ‘adaptable commitment interface,’ the SwitchTube mobile app, in which users can toggle between two interfaces when watching YouTube videos: Focus Mode (search-first) and Explore Mode (recommendations-first). In a three-week field deployment with 46 US participants, we evaluate how the ability to switch between interfaces affects user experience, finding that it provides users with a greater sense of agency, satisfaction, and goal alignment. We conclude with design implications for how adaptable commitment interfaces can support digital wellbeing.
In this study, we propose a method to predict whether a web searcher will purchase a camera in a near future based on his/her web search log. With the increasing popularity of online shopping at EC sites, more and more users are searching for products through web searches and actually purchasing them at EC sites. This indicates that, by analyzing the query log of a searcher, it is possible to predict whether the searcher will purchase the product in the near future. Therefore, we construct a classifier by collecting past web search query logs of searchers who have purchased cameras and those who have not purchased them. In the experiment, we used a web search query log of Yahoo! JAPAN and the product purchase histories of Yahoo! JAPAN Shopping to verify the results. We collected thousands of users who purchased cameras in a certain period and other users in the same number who didn’t purchase but issued queries related to cameras. By analyzing the classifier trained with the prepared dataset, we verify the accuracy of the prediction, the period of time required for the prediction, and whether there are any characteristic words that suggest the purchase.
Microblog activity logs are useful to determine user’s interest and sentiment towards specific and broader category of events such as natural disaster and national election. In this paper, we present a corpus model to show how personal attitudes can be predicted from social media or microblog activities for a specific domain of events such as natural disasters. More specifically, given a user’s tweet and an event, the model is used to predict whether the user will be willing to help or show a positive attitude towards that event or similar events in the future. We present a new dataset related to a specific natural disaster event, i.e. Hurricane Harvey, that distinguishes user’s tweets into positive and non-positive attitudes. We build Term Embeddings for Tweet (TEmT) to generate features to model personal attitudes for arbitrary user’s tweets. In addition, we present sentiment analysis on the same disaster event dataset using enhanced feature learning on TEmT generated features by applying Convolutional Neural Network (CNN). Finally, we evaluate the effectiveness of our method by employing multiple classification techniques and comparative methods on the newly created dataset.
Eliciting user preferences from purchase records for performing purchase prediction is challenging because negative feedback is not explicitly observed, and because treating all non-purchased items equally as negative feedback is unrealistic. Therefore, in this study, we present a framework that leverages the past click records of users to compensate for the missing user–item interactions of purchase records, i.e., non-purchased items. We begin by formulating various model assumptions, each one assuming a different order of user preferences among purchased, clicked-but- not-purchased, and non-clicked items, to study the usefulness of leveraging click records. We implement the model assumptions using the Bayesian personalized ranking model, which maximizes the area under the curve for bipartite ranking. However, we argue that using click records for bipartite ranking needs a meticulously designed model because of the relative unreliableness of click records compared with that of purchase records. Therefore, we ultimately propose a novel learning-to-rank method, called P3Stop, for performing purchase prediction. The proposed model is customized to be robust to relatively unreliable click records by particularly focusing on the accuracy of top-ranked items. Experimental results on two real-world e-commerce datasets demonstrate that P3STop considerably outperforms the state-of-the-art implicit-feedback-based recommendation methods, especially for top-ranked items.
Users in e-commerce tend to click on items of their interest. Eventually, the more frequently an item is clicked by a user, the more likely the item will be purchased by the user after all. However, what if a user clicked on every item only once before purchases? This is a frequently observed user behavior in reality, but predicting which of the clicked items will be purchased is a challenging task. This paper addresses a practical yet widely overlooked task of predicting purchase items within a non-duplicate click session, i.e., a session in which every item is clicked only once. We propose an encoder–decoder neural architecture to simultaneously model users’ click and purchase behaviors. The encoder captures a user’s intent contained in the user’s click session, and the decoder, which is equipped with pointer network via a switch gate, extracts relevant clicked items for future purchase candidates. To the best of our knowledge, our work is the first to address the task of purchase prediction given non-duplicate click sessions. Experiments demonstrate that our proposed method outperforms the state-of-the-art purchase prediction methods by up to 18% in terms of recall.
Online production communities such as Wikipedia and OpenStreetMap play an important role in connecting the public with major events in society. The popularity of a major event, together with the popularity of online communities brings the general public to collaborate on and co-create knowledge about the event. The high level of interest in capturing what draws the attention of society can particularly help online production communities meet one of the essential challenges they face: attracting and retaining newcomers. In this work, we explore how newcomers in such communities respond to knowledge production around major societal events. Analysis of the participation of 506 newcomers to Wikipedia articles related to three highly popular events shows that the popularity of the events attracts a new wave of users to the online community. These newcomers provide valuable contributions to the community, however, at a differing level depending on their initial motivation and experiences. Those participants who joined the online community solely to contribute to one topic or event are more likely to face challenges in contribution and leave Wikipedia after limited contribution. We discuss factors and patterns of newcomers’ early and longer-term participation in Wikipedia in relation to three popular events.
What are the intents or goals behind human interactions with image search engines? Knowing why people search for images is of major concern to Web image search engines because user satisfaction may vary as intent varies. Previous analyses of image search behavior have mostly been query-based, focusing on what images people search for, rather than intent-based, that is, why people search for images. To date, there is no thorough investigation of how different image search intents affect users» search behavior. In this paper, we address the following questions: (1)Why do people search for images in text-based Web image search systems? (2)How does image search behavior change with user intent? (3)Can we predict user intent effectively from interactions during the early stages of a search session? To this end, we conduct both a lab-based user study and a commercial search log analysis. We show that user intents in image search can be grouped into three classes: Explore/Learn, Entertain, and Locate/Acquire. Our lab-based user study reveals different user behavior patterns under these three intents, such as first click time, query reformulation, dwell time and mouse movement on the result page. Based on user interaction features during the early stages of an image search session, that is, before mouse scroll, we develop an intent classifier that is able to achieve promising results for classifying intents into our three intent classes. Given that all features can be obtained online and unobtrusively, the predicted intents can provide guidance for choosing ranking methods immediately after scrolling.
As online shopping becomes increasingly popular, users perform more product search to purchase items. Previous studies have investigated people's online shopping behaviours and ways to predict online purchases. However, from a user perspective, there still lacks an in-depth understanding of why users search, how they interact with, and perceive the product search results. In this paper, we conduct both a user study and a log analysis to we address the following three questions: (1) what are the intents of users underlying their search activities? (2) do users behave differently under different search intents? and (3) how does user perceived satisfaction relate to their search behaviour as well as search intents, and can we predict product search satisfaction with interaction signals? Based on an online survey and search logs collected from a major commercial product search engine, we show that user intents in product search fall into three categories: Target Finding (TF), Decision Making (DM) and Exploration (EP). Through a log analysis and a user study, we observe different user interaction patterns as well as perceived satisfaction under these three intents. Using a series of user interaction features, we demonstrate that we can effectively predict user satisfaction, especially for TF and DM intents.
There appears to be general agreement among social psychologists that most human behavior is goal-directed (e. g., Heider, 1958 ; Lewin, 1951). Being neither capricious nor frivolous, human social behavior can best be described as following along lines of more or less well-formulated plans. Before attending a concert, for example, a person may extend an invitation to a date, purchase tickets, change into proper attire, call a cab, collect the date, and proceed to the concert hall. Most, if not all, of these activities will have been designed in advance; their execution occurs as the plan unfolds. To be sure, a certain sequence of actions can become so habitual or routine that it is performed almost automatically, as in the case of driving from home to work or playing the piano. Highly developed skills of this kind typically no longer require conscious formulation of a behavioral plan. Nevertheless, at least in general outline, we are normally well aware of the actions required to attain a certain goal. Consider such a relatively routine behavior as typing a letter. When setting this activity as a goal, we anticipate the need to locate a typewriter, insert a sheet of paper, adjust the margins, formulate words and sentences, strike the appropriate keys, and so forth. Some parts of the plan are more routine, and require less conscious thought than others, but without an explicit or implicit plan to guide the required sequence of acts, no letter would get typed.
Consumers shop online for both goal-oriented and experiential reasons. However, goal-oriented motives are more common among online shoppers than are experiential motives. This article identifies and discusses attributes that facilitate goal-oriented online shopping, including accessibility/convenience, selection, information availability, and lack of unwanted sociality from retail sales help or shopping partners such as spouses. Importantly, consumers report that shopping online results in a substantially increased sense of freedom and control as compared to offline shopping. While consumers are more likely to describe offline rather than online shopping in experiential terms, evidence of experiential motivations for online shopping is emerging. Also, while closing transactions at web sites is one important e-commerce goal, companies should not lose site of the continuing importance and power of their web site as an information and communications vehicle.
The causal role of students' self-efficacy beliefs and academic goals in self-motivated academic attainment was studied using path analysis procedures. Parental goal setting and students' self-efficacy and personal goals at the beginning of the semester served as predictors of students' final course grades in social studies. In addition, their grades in a prior course in social studies were included in the analyses. A path model of four self-motivation variables and prior grades predicted students 'final grades in social studies, R = .56. Students' beliefs in their efficacy for self-regulated learning affected their perceived self-efficacy for academic achievement, which in turn influenced the academic goals they set for themselves and their final academic achievement. Students' prior grades were predictive of their parents' grade goals for them, which in turn were linked to the grade goals students set for themselves. These findings were interpreted in terms of the social cognitive theory of academic self-motivation.
Tested the prediction that feedback and goals would be interactively related to performance. This prediction complements the findings of Locke that knowledge alone is not a sufficient condition for effective performance. It was also hypothesized that feedback would facilitate the display of individual differences in goal setting and hence the goal setting–performance relationship. 38 undergraduates in a feedback group and 48 in a no-feedback group were evaluated on a number comparison task. Results support the hypotheses by indicating that (a) the individual differences in self-set goals were significantly higher in the feedback group than in the no-feedback group, and that (b) the relationship between goals and performance was significantly higher in the feedback condition than in the no-feedback condition. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Investigated the effects of goal setting by supervisors on 4 performance criteria. A factor analysis of a questionnaire administered to 292 pulpwood producers indicated that goal setting is correlated with high productivity and a low number of injuries only when it is accompanied by supervision. Goal setting without immediate supervision was related to employee turnover. Supervision alone did not correlate with any performance criterion. No relationship was found between the 2 job behaviors and absenteeism. The factor analysis results pertaining to productivity were partially corroborated in a 2nd study involving data collected from 892 additional producers. (24 ref) (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Goals and related constructs are ubiquitous in psychological research and span the history of psychology. Research on goals has accumulated sporadically through research programs in cognition, personality, and motivation. Goals are defined as internal representations of desired states. In this article, the authors review the theoretical development of the structure and properties of goals, goal establishment and striving processes, and goal-content taxonomies. They discuss affect as antecedent, consequence, and content of goals and argue for integrating across psychological content areas to study goal-directed cognition and action more efficiently. They emphasize the structural and dynamic aspects of pursuing multiple goals, parallel processing, and the parsimony provided by the goal construct. Finally, they advocate construct validation of a taxonomy of goals. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
to establish a general theory of work behavior, one must begin with the concept of action / action is goal-oriented behavior that is organized in specific ways by goals, information integration, plans, and feedback and can be regulated consciously or via routines / describe general [German] theory along these lines / this is quite a different theory from the typical American theories in industrial and organizational psychology (PsycINFO Database Record (c) 2012 APA, all rights reserved)
The purpose of this article is to examine the role of goal commitment in goal-setting research. Despite Locke's (1968) specification that commitment to goals is a necessary condition for the effectiveness of goal setting, a majority of studies in this area have ignored goal commitment. In addition, results of studies that have examined the effects of goal commitment were typically inconsistent with conceptualization of commitment as a moderator. Building on past research, we have developed a model of the goal commitment process and then used it to reinterpret past goal-setting research. We show that the widely varying sizes of the effect of goal difficulty, conditional effects of goal difficulty, and inconsistent results with variables such as participation can largely be traced to main and interactive effects of the variables specified by the model. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Attempted a field replication of J. F. Bryan and E. A. Locke's (see record
1967-15919-001) general version of Parkinson's Law (i.e., effort is adjusted to the difficulty of the task). The principle implies greater effort on a given task when time restrictions are placed on task completion than when there are no such restrictions. As predicted, members of logging crews (N = 379) showed a significantly higher rate of output when limited to mill deliveries on 1 or 2 days/wk than when no such buying restrictions were in effect. Since both owners and crew members were paid on a piece-rate basis, there was an incentive to maximize production early when restrictions were operative. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Knowledge workers spend the majority of their working hours processing and manipulating information. These users face continual costs as they switch between tasks to retrieve and create information. The TaskTracer project at Oregon State University is investigating the possibilities of a desktop software system that will record in detail how knowledge workers complete tasks, and intelligently leverage that information to increase efficiency and productivity. Our approach combines human-computer interaction and machine learning to assign each observed action (opening a file, saving a file, sending an email, cutting and pasting information, etc.) to a task for which it is likely being performed. In this paper we report on ways we have applied machine learning in this environment and lessons learned so far.
Accurately capturing user preferences over time is a great practical challenge in recommender systems. Simple correlation over time is typically not meaningful, since users change their preferences due to different external events. User behavior can often be determined by individual's long-term and short-term preferences. How to represent users' long-term and short-term preferences? How to leverage them for temporal recommendation? To address these challenges, we propose Session-based Temporal Graph (STG) which simultaneously models users' long-term and short-term preferences over time. Based on the STG model framework, we propose a novel recommendation algorithm Injected Preference Fusion (IPF) and extend the personalized Random Walk for temporal recommendation. Finally, we evaluate the effectiveness of our method using two real datasets on citations and social bookmarking, in which our proposed method IPF gives 15%-34% improvement over the previous state-of-the-art.
This paper investigates the uses of social networking site Facebook, and the gratifications users derive from those uses. In the first study, 137 users generated words or phrases to describe how they used Facebook, and what they enjoyed about their use. These phrases were coded into 46 items which were completed by 241 Facebook users in Study 2. Factor analysis identified seven unique uses and gratifications: social connection, shared identities, content, social investigation, social network surfing and status updating. User demographics, site visit patterns and the use of privacy settings were associated with different uses and gratifications.
Classic IR (information retrieval) is inherently predicated on users searching for information, the so-called "information need". But the need behind a web search is often not informational -- it might be navigational (give me the url of the site I want to reach) or transactional (show me sites where I can perform a certain transaction, e.g. shop, download a file, or find a map). We explore this taxonomy of web searches and discuss how global search engines evolved to deal with web-specific needs.
The benefits of research tools in exploratory search with user interface capability are discussed. Exploratoty search provide information along with pitfalls and costs and helps in gaining complete knowledge about the product. The tools to support exploratory searching make possible to use use web resources to conduct lookup, learning, and investigate tasks with fluid user interfaces. A good web design supports exploratory search together with client-side applications, that provides people with ways to trade off personal behavior data for added value services. It is recommended that more computational resources should be designed for exploratory search, that will provide easy to apply exploratory search tools and help information seekers in the use of information resources.
Several opinion polls have found that many consumers resist making purchases via the Internet because of their concerns about the privacy of the personal information they provide to Internet merchants. Using the theory of planned behavior as its basis, this study investigated the relationships among beliefs about Internet privacy and trustworthiness, along with beliefs about perceived behavioral control and the expectations of important others, and online purchasing behavior. Data were collected from 193 college students. Analysis of the data indicates that beliefs about trustworthiness positively affect attitudes toward buying online, which in turn positively affect purchasing behavior. Beliefs about self-efficacy regarding purchasing positively affect perceived behavioral control, which in turn affects online purchasing behavior. In short, respondents who believed in the trustworthiness of the Internet and in their own abilities to buy online were more likely to make Internet purchases than were those without such beliefs.
Past research has established that, while self-reports of purchase intentions can predict behavior, various factors affect the strength of the intentions-behavior link. This article explores one such factor: the impact of merely measuring intent. Our specific question concerns the impact of measuring intent on subsequent purchase behavior. Prior research suggests a mere-measurement hypothesis: that merely measuring intent will increase subsequent purchase behavior. We also suggest a polarization hypothesis: that repeated intent questions will have a polarizing effect on behavior. The results reveal that the effect of merely asking intent to buy once is an increase in the subsequent purchase rate. The effect of repeatedly asking intent for those with low levels of intent is a decreased propensity to buy with repeated measurements. These two effects are reduced given prior experience with the product. The implications of these findings and opportunities for future research are discussed.
This study investigated the interactive effects of achievement orientation and evaluative focus of assigned, task-specific goals on intrinsic motivation for an enjoyable pinball game. Regression analyses revealed that the effect of performance or mastery-focused goals on intrinsic motivation varied as a function of achievement orientation. Specifically, the provision of task-specific standards with a performance focus enhanced intrinsic motivation for achievement-oriented individuals, whereas the assignment of such goals proved deleterious to the intrinsic motivation of those low in achievement orientation. Individuals low in achievement orientation displayed the highest levels of intrinsic motivation when provided with mastery-focused goals. A similar pattern of effects was obtained on competence valuation and task involvement, both of which were additionally validated as mediators of the direct effects on intrinsic motivation.
On the basis of postulates derived from socioemotional selectivity theory, the authors explored the extent to which future time perspective (FTP) is related to social motivation, and to the composition and perceived quality of personal networks. Four hundred eighty German participants with ages ranging from 20 to 90 years took part in the study. In 2 card-sort tasks, participants indicated their partner preference and goal priority. Participants also completed questionnaires on personal networks and social satisfaction. Older people, as a group, perceived their future time as more limited than younger people. Individuals who perceived future time as being limited prioritized emotionally meaningful goals (e.g., generativity, emotion regulation), whereas individuals who perceived their futures as open-ended prioritized instrumental or knowledge-related goals. Priority of goal domains was found to be differently associated with the size, composition, and perceived quality of personal networks depending on FTP. Prioritizing emotion-regulatory goals was associated with greater social satisfaction and less perceived strain with others when participants perceived their future as limited. Findings underscore the importance of FTP in the self-regulation of social relationships and the subjective experience associated with them.
The authors investigate whether the use of segmentation can improve the accuracy of sales forecasts based on stated purchase intent. The common current practice is to prepare a sales forecast by using purchase intent and observed historical patterns in purchase rates given level of intent. The authors show that the accuracy of sales forecasts based on purchase intent can be improved by first using certain kinds of segmentation methods to segment the panel members. The main empirical finding is that more accurate sales forecasts appear to be obtained by applying statistical segmentation methods that distinguish between dependent and independent variables (e.g., CART, discriminant analysis) than by applying simpler direct clustering approaches (e.g., a priori segmentation or K-means clustering). The results further reveal that meaningful segments are present and identifiable that vary in their subsequent purchase rates for a given level of intent. This identification has important implications for areas such as target marketing, as it indicates which customer segments will actually fulfill their intentions. One key substantive finding is that households in a (demographic/product-usage-based) segment having an a priori high propensity to purchase are also more likely to fulfill a positive intention to purchase.
Online e-commerce applications are becoming a primary vehicle for people to find, compare, and ultimately purchase products. One of the fundamental questions that arises in e-commerce is to characterize, understand, and model user long-term purchasing intent, which is important as it allows for personalized and context relevant e-commerce services.
In this paper we study user activity and purchasing behavior with the goal of building models of time-varying user purchasing intent. We analyze the purchasing behavior of nearly three million Pinterest users to determine short-term and long-term signals in user behavior that indicate higher purchase intent.
We find that users with long-term purchasing intent tend to save and clickthrough on more content. However, as users approach the time of purchase their activity becomes more topically focused and actions shift from saves to searches. We further find that purchase signals in online behavior can exist weeks before a purchase is made and can also be traced across different purchase categories. Finally, we synthesize these insights in predictive models of user purchasing intent. Taken together, our work identifies a set of general principles and signals that can be used to model user purchasing intent across many content discovery applications.
The authors investigate whether the use of segmentation can improve the accuracy of sales forecasts based on stated purchase intent. The common current practice is to prepare a sales forecast by using purchase intent and observed historical patterns in purchase rates given level of intent. The authors show that the accuracy of sales forecasts based on purchase intent can be improved by first using certain kinds of segmentation methods to segment the panel members. The main empirical finding is that more accurate sales forecasts appear to be obtained by applying statistical segmentation methods that distinguish between dependent and independent variables (e.g., CART, discriminant analysis) than by applying simpler direct clustering approaches (e.g., a priori segmentation or K-means clustering). The results further reveal that meaningful segments are present and identifiable that vary in their subsequent purchase rates for a given level of intent. This identification has important implications for areas such as target marketing, as it indicates which customer segments will actually fulfill their intentions. One key substantive finding is that households in a (demographic/product-usage-based) segment having an a priori high propensity to purchase are also more likely to fulfill a positive intention to purchase.
We develop new understanding of how people engage in digital curation. We interview twenty users of Pinterest, a social curation platform. We find that through collecting, organizing, and sharing image bookmarks, users engage in processes of everyday ideation. That is, they use digital found objects as creative resources to develop ideas for shaping their lives. Curators assemble information into new contexts, forming and sharing ideas with practical and emotional value. We investigate cognitive and social aspects of creativity that affect the digital curation practices of everyday ideation. We derive implications for the design of curation environments that support information-based ideation.
Pinterest is a popular social curation site where people collect, organize, and share pictures of items. We studied a fundamental issue for such sites: what patterns of activity attract attention (audience and content reposting)-- We organized our studies around two key factors: the extent to which users specialize in particular topics, and homophily among users. We also considered the existence of differences between female and male users. We found: (a) women and men differed in the types of content they collected and the degree to which they specialized; male Pinterest users were not particularly interested in stereotypically male topics; (b) sharing diverse types of content increases your following, but only up to a certain point; (c) homophily drives repinning: people repin content from other users who share their interests; homophily also affects following, but to a lesser extent. Our findings suggest strategies both for users (e.g., strategies to attract an audience) and maintainers (e.g., content recommendation methods) of social curation sites.
Two experiments were designed to test the effects of different time limits on time taken to complete a task (Parkinson's law) and to determine whether goal-setting mediated the effects of time limits on performance rate. It was hypothesized that the different time limits would produce differences in performance rate only if and to the degree that different goals were set by Ss in the different conditions. Two groups were run in each experiment. On the first three trials, the Excess-time Ss were given twice as much time as was needed to complete an addition task, and the Minimum-time Ss were given just enough time to finish it. On the fourth trial, the Ss in both groups were allowed to work at their own pace, and on the fifth trial the Ss in both groups were asked to work as fast as their estimated “fastest possible” time. Both groups in both experiments performed in accordance with Parkinson's Law on trials 1 to 3 (the Excess Ss taking longer to complete the task), and the hypothesis that goal setting mediated the Parkinson effect was supported. It was also found, however, that the same goal could mean different things in different contexts. For instance, the Ss in the Excess group who were trying to work “quickly but not as fast as possible” on trials 1 to 3 worked at a significantly slower pace on these trials than the Minimum Ss who said they were working for the same goal during this period. Although the Parkinson effect did not extend to trials 4 and 5, goals were related to performance within groups on these trials.
A proposed theory of planned behavior, an extension of Ajzen and Fishbein's (1980, Understanding attitudes and predicting social behavior. Englewood-Cliffs, NJ: Prentice-Hall) theory of reasoned action, was tested in two experiments. The extended theory incorporates perceived control over behavioral achievement as a determinant of intention (Version 1) as well as behavior (Version 2). In Experiment 1, college students' attendance of class lectures was recorded over a 6-week period; in Experiment 2, the behavioral goal was getting an “A” in a course. Attitudes, subjective norms, perceived behavioral control, and intentions were assessed halfway through the period of observation in the first experiment, and at two points in time in the second experiment. The results were evaluated by means of hierarchical regression analyses. As expected, the theory of planned behavior permitted more accurate prediction of intentions and goal attainment than did the theory of reasoned action. In both experiments, perceived behavioral control added significantly to the prediction of intentions. Its contribution to the prediction of behavior was significant in the second wave of Experiment 2, at which time the students' perceptions of behavioral control had become quite accurate. Contrary to expectations, there was little evidence for interactions between perceived behavioral control and the theory's other independent variables.
Recently, it has been proposed that creating compelling experiences in the distinctive consumption environment defined by the Internet depends on facilitating a state of flow. Although it has been established that consumers do, in fact, experience flow while using the Web, consumer researchers do not as yet have a comprehensive understanding of the specific activities during which consumers actually have these experiences.
One fruitful focus of research on online consumer experience has been on two distinct categories of consumption behavior—goal directed and experiential consumption behavior. Drawing distinctions between these behaviors for the Web may be particularly important because the experiential process is, for many individuals, as or even more important than the final instrumental result. However, the general and broad nature of flow measurement to date has precluded a precise investigation of flow during goal-directed versus experiential activities.
In this article, we explore this issue, investigating whether flow occurs during both experiential and goal-directed activities, if experiential and goal-directed flow states differ in terms of underlying constructs, and what the key characteristics are—based on prior theory—that define “types” of flow experiences reported on the Web. Our approach is to perform a series of quantitative analyses of qualitative descriptions of flow experiences provided by Web users collected in conjunction with the 10th GVU WWW User Survey. In contrast with previous research that suggests flow would be more likely to occur during recreational activities than task-oriented activities, we found more evidence of flow for task-oriented rather than experiential activities, although there is evidence flow occurs under both scenarios. As a final note, we argue that the role that goal-directed and experiential activities may play in facilitating the creation of compelling online environments may also be important in a broader consumer policy context.
Two studies are reported which deal with the directing function of goals in task performance. Subjects were given, on each trial, knowledge of their scores on two or more independent (uncorrelated) performance dimensions; but they set goals to maximize (or minimize) their scores on only one performance dimension. It was predicted that scores on a given dimension would be higher (or lower) when the subject was trying to maximize (or minimize) his score on that dimension than when he was trying for some other goal. The predictions were generally confirmed.
This paper summarizes and integrates research concerned with a long-neglected topic in psychology: the relationship between conscious goals and intentions and task performance. The basic promise of this research is that an individual's conscious ideas regulate his actions. Studies are cited demonstrating that: (1) hard goals produce a higher level of performance (output) than easy goals; (2) specific hard goals produce a higher level of output than a goal of “do your best”; and (3) behavioral intentions regulate choice behavior. The theory also views goals and intentions as mediators of the effects of incentives on task performance. Evidence is presented supporting the view that monetary incentives, time limits, and knowledge of results do not affect performance level independently of the individual's goals and intentions. A theoretical analysis supports the same view with respect to three other incentives: participation, competition, and praise and reproof. Finally, behavioral intentions were found to mediate the effects of money and “verbal reinforcement” on choice behavior. It is concluded that any adequate theory of task motivation must take account of the individual's conscious goals and intentions. The applied implications of the theory are discussed.
Notes that meta-analyses of research using the theory of reasoned action and the theory of planned behavior show that these models explain on average between 40% and 50% of the variance in intention, and between 19% and 38% of the variance in behavior. This paper evaluates the performance of these models in predicting and explaining intentions and behavior. The distinction between prediction and explanation, the different standards of comparison against which predictive performance can be judged, and the use of percentage of variance explained as a measure of effect size are discussed. Reasons why the models do not always predict as well expected are presented. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
In this paper we examine goal setting and feedback as possible sources of perceived organizational support (Eisenberger, Huntington, Hutchison, & Sowa, 1986). Surveys were administered to 337 employees from 3 different organizations in the southern California area. Two path models were examined: one that allowed goal setting and feedback to have direct effects on perceived support and organizational commitment, and one that only allowed goal setting and feedback to have direct effects on perceived support. Results show that goal setting and feedback do not have direct effects on organizational commitment but, rather, affect commitment indirectly by way of their effects on employee perceptions of support from the organization.
The social networking site, Facebook, has gained an enormous amount of popularity. In this article, we review the literature on the factors contributing to Facebook use. We propose a model suggesting that Facebook use is motivated by two primary needs: (1) The need to belong and (2) the need for self-presentation. Demographic and cultural factors contribute to the need to belong, whereas neuroticism, narcissism, shyness, self-esteem and self-worth contribute to the need for self presentation. Areas for future research are discussed.
This study examines the influence of demographic variables and dimensions of motivational factors of two types of consumer Web use: percentage of weekly Web surfing time spent searching for product and service-related information and online shopping and transactions. It combines data from two sources: a self-administered survey of 59 undergraduates in an introductory communication course at Cornell University; and a mail/Web survey of 59 New York State residents who had reported subscribing to an online service in a previous mail survey. We found distinctively different patterns of relationships among demographics and motivational factors for the two types of dependent variables. Most importantly, transactional privacy concerns were found to be negatively related to percentage of time spent on product searches and online shopping, while economic motivations had a positive influence. In addition, online shopping was found to be predicted by information motivations, interactive control motivations, and socialization motivations. Implications for Web-based commerce and advertising are discussed.
The future time perspective is described first in its relationship to learning and to motivation. The cognitive elaboration of human needs in plans, projects, and tasks is emphasized, and it is shown that the future time perspective in men is to be studied with regard to these plans, aspirations, and tasks.The results of two research programs are summarized: 1) The depth of time perspective in which the motivational objects of different categories of subjects are located is investigated in a cross-cultural study. 2) In a series of experiments on learning the influence of a future time perspective with regard to the response given is investigated in “open” and “closed” tasks. On the basis of these results some hypotheses on learning and behavior with regard to motivational and cognitive processes are proposed. Learning is conceived of as a process by which behavioral responses are incorporated or embedded in the dynamic systems of the individual. The arousal of the dynamic system activates at the same time the behavioral pattern embedded in it, and in this way the gap between learning and action or performance can be bridged. As to the influence of cognition on behavior it is emphasized that cognitive content is a kind of “precipitate” or cristalization of earlier behavioral contacts and that its influence on actual behavior is, therefore, no separate problem.
Numerous studies of goal setting have found that specific and difficult or high-level goals lead to higher performance than vague (do best), easy goals, or no goals. However, often it has been asserted in the literature that specific goals as such lead to higher performance than vague goals even though goal theory makes no such claim. However, no previous study (with one partial exception) has actually separated the effects of goal level from those of goal specificity. It was predicted that when the two goal attributes were separated, goal level would affect level of performance whereas goal specificity would affect the variability of performance. Two experiments were conducted to test these hypotheses. The first used a reaction time task and the second an idea-generation task. The results of both studies supported the hypotheses. However, one of the two present studies and two previous studies found that high goal levels can also affect performance variance.
Recent advances in search users' click modeling consider both users' search queries and click/skip behavior on documents to infer the user's perceived relevance. Most of these models, including dynamic Bayesian networks (DBN) and user browsing models (UBM), use probabilistic models to understand user click behavior based on individual queries. The user behavior is more complex when her actions to satisfy her information needs form a search session, which may include multiple queries and subsequent click behaviors on various items on search result pages. Previous research is limited to treating each query within a search session in isolation, without paying attention to their dynamic interactions with other queries in a search session.
Investigating this problem, we consider the sequence of queries and their clicks in a search session as a task and propose a task-centric click model~(TCM). TCM characterizes user behavior related to a task as a collective whole. Specifically, we identify and consider two new biases in TCM as the basis for user modeling. The first indicates that users tend to express their information needs incrementally in a task, and thus perform more clicks as their needs become clearer. The other illustrates that users tend to click fresh documents that are not included in the results of previous queries. Using these biases, TCM is more accurately able to capture user search behavior. Extensive experimental results demonstrate that by considering all the task information collectively, TCM can better interpret user click behavior and achieve significant improvements in terms of ranking metrics of NDCG and perplexity.
In most previous work on personalized search algorithms, the results for all queries are personalized in the same manner. However, as we show in this paper, there is a lot of variation across queries in the benefits that can be achieved through personalization. For some queries, everyone who issues the query is looking for the same thing. For other queries, different people want very different results even though they express their need in the same way. We examine variability in user intent using both explicit relevance judgments and large-scale log analysis of user behavior patterns. While variation in user behavior is correlated with variation in explicit relevance judgments the same query, there are many other factors, such as result entropy, result quality, and task that can also affect the variation in behavior. We characterize queries using a variety of features of the query, the results returned for the query, and people's interaction history with the query. Using these features we build predictive models to identify queries that can benefit from personalization. Categories and Subject Descriptors
Different from past online shopping research which focuses on exploring the cognitive and attitudinal aspects of online consumption behaviors, this study focuses on examining how online shopping motivation and product type affect the searching and spending behavior in an online website. In particular, the interactive effect between online shopping motivation and product type are explored in the study. A 2×2 between-subjects factorial experiment with two shopping motivations (i.e., goal-oriented and experiential) and two product types (i.e., hedonic and utilitarian product) was conducted. In addition to the main effects of online shopping motivation and product type, the results show that there is an interactive effect between online shopping motivation and product type on budget control. Participants conducted the loosest budget control when purchasing hedonic products in a goal-oriented online shopping trip because the shopping motivation provides a good justification for online shoppers to purchase hedonic goods without guilt or regret.
Goal-directed and experiential online shopping are two general categories of consumer behavior on the Internet. Internet stores promoting online transactions may face challenges regarding how to satisfy distinct online shoppers. This study develops a dual-state of cognitive and affective reactions to online services and flow experiences to predict goal-directed and experiential online consumer behavior. This survey mailed 300 questionnaires to a random sample of volunteers with substantial online shopping experience. Of these, 150 samples were returned for data analysis. Empirical results from the survey indicate that the dual-state of online services is more appropriate for assessing goal-directed purchase intentions than for examining the willingness to purchase experientially. In contrast, the dual-state of flow experiences is better at assessing goal-directed purchase intentions and the willingness to purchase experientially. Finally, this study features implications for academics and practitioners.
Search engine advertising has become a significant element of the Web browsing experience. Choosing the right ads for the query and the order in which they are displayed greatly affects the prob- ability that a user will see and click on each ad. This ranking has a strong impact on the revenue the search engine receives from the ads. Further, showing the user an ad that they prefer to click on improves user satisfaction. For these reasons, it is important to be able to accurately estimate the click-through rate of ads in the system. For ads that have been displayed repeatedly, this is empir- ically measurable, but for new ads, other means must be used. We show that we can use features of ads, terms, and advertisers to learn a model that accurately predicts the click-though rate for new ads. We also show that using our model improves the con- vergence and performance of an advertising system. As a result, our model increases both revenue and user satisfaction. Categories and Subject Descriptors
To study the processes by which past behavior influences future behavior, participants were led to believe that without being aware of it, they had expressed either support for or opposition to the institution of comprehensive exams. Judgment and response time data suggested that participants' perceptions of their past behavior often influenced their decisions to repeat the behavior. This influence was partly the result of cognitive activity that influenced participants' cognitions about specific behavioral consequences and the attitude they based on these cognitions. More generally, however, feedback about past behavior had a direct effect on participants' attitudes and ultimate behavioral decisions that was independent of the outcome-specific cognitions. Results are discussed in terms of their implications for biased scanning of memory, dissonance reduction, self-perception, and the use of behavior as a heuristic.
The subjective sense of future time plays an essential role in human motivation. Gradually, time left becomes a better predictor
than chronological age for a range of cognitive, emotional, and motivational variables. Socioemotional selectivity theory
maintains that constraints on time horizons shift motivational priorities in such a way that the regulation of emotional states
becomes more important than other types of goals. This motivational shift occurs with age but also appears in other contexts
(for example, geographical relocations, illnesses, and war) that limit subjective future time.
This paper discusses the discounted utility (DU) model: its historical development, underlying assumptions, and "anomalies"--the empirical regularities that are inconsistent with its theoretical predictions. We then summarize the alternate theoretical formulations that have been advanced to address these anomalies. We also review three decades of empirical research on intertemporal choice, and discuss reasons for the spectacular variation in implicit discount rates across studies. Throughout the paper, we stress the importance of distinguishing time preference, per se, from many other considerations that also influence intertemporal choices.
A meta-analytic study was conducted involving primarily published research from 1966 to 1984 and focusing on the relationship between goal-setting variables and task performance. Two major sets of studies were analyzed, those contrasting hard goals (goal difficulty) versus easy goals, and those comparing specific hard goals (goal specificity/difficulty) versus general goals, “do best” instructions, or no goal. As expected, strong support was obtained for the goal difficulty and goal specificity/difficulty components of E. A. Locke's (1968a, Organizational Behavior and Human Performance, 3, 157–189) theory. A two-stage approach was employed to identify potential moderators of the goal difficulty and goal specificity/difficulty—performance relationships. Setting (laboratory versus field) was identified as a moderator of the relationship between goal specificity/difficulty and task performance. Two supplemental meta-analyses yielded support for the efficacy of combining specific hard goals with feedback versus specific hard goals without feedback and for participatively set goals versus assigned goal setting (when goal level is held constant), although this latter finding was interpreted as inconclusive based on the limited studies available. Implications for future research are addressed.
150 million people finding ideas on pinterest
- B Silbermann
User intent affects filter design
- K Sherwin
Exploratory search: from finding to understanding. CACM
- G Marchionini