No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
A context-aware, info-bead and fuzzy inference approach to notification management
... The work of Reference [32] is preliminary and attempts only to predict which device should be selected to deliver the notification, however, the set of notification data used to train the algorithms and evaluate the system result is partially synthetic and assumes that the data available for the notification is explicit. In contrast, the management system proposed by Reference [33], attempts to predict the most opportune moment for the notification to be delivered to the user using a set of data obtained from real users. Only the abstract derivation is used by the manager to predict delivery time. ...
... 1. Reference; 2. Solution Name and Year; 3. If the implementation makes use of the user's location; 4. Whether the implementation determines privacy preferences or implements actions to maintain private data; 5. Compares whether the notification solution makes use of multiple devices, being more than one mobile device or another smart device; 6. Informs whether the proposed solution is applied directly to the user's mobile device through an application; 7. Informs whether the solution uses human participation to verify the relevance and manual adjustments in the context of notification management. [27] Attelia (2014) X X [28] Message Monitor (2014) X X X X [29] Face-to-Face (2014) X X X [30] Desktop Notifications (2014) X X X X [31] Intelligent Push (2015) X X [32] Notification Collector (2015) X X X [33] NAbsMobile (2016) X X X X [34] No name (2017) X X X [19] Smartnotify (2018) X X [35] Notification Log (2018) X X X X This Work PRISER (2019) X X X X Although related work addresses location and notification management, many of these studies do not address the privacy features of the user's environment. Therefore, the present work proposes a module of notifications, to make the environment informative and dynamic to control privacy parameters. ...
... However, from a high-level perspective, each component of this taxonomy was based on a series of related works that helped us identify the needs of the NMS. The works used are represented by Table 2. [12,19,33] In this section, we discuss some enabling technologies, which compose the proposed taxonomic model. The taxonomy serves to classify rules and parameters, resulting in a better understanding of the functionality [51]. ...
With the growing number of mobile devices receiving daily notifications, it is necessary to manage the variety of information produced. New smart devices are developed every day with the ability to generate, send, and display messages about their status, data, and information about other devices. Consequently, the number of notifications received by a user is increasing and their tolerance may decrease in a short time. With this, it is necessary to develop a management system and notification controls. In this context, this work proposes a notification and alert management system called PRISER. Its focus is on user profiles and environments, applying data privacy criteria.
... In this paper, we evaluate a modest notification dataset [16] captured in-the-wild and comprised of: ...
... For privacy reasons the content of notifications were dismissed however. In contrast, Fraser et al. [16] do not completely discard the notification content of their in-the-wild notifications but instead enable the user to uplift the sensitive information to an abstract form. If notification content could be generated synthetically while maintaining real-world underlying traits, there would be no need to abstract or discard valuable contextual information such as this. ...
... In a recent study of notification management [16] two notification datasets were harvested from two users'. An Android device was used to capture features of incoming mobile notifications via an application developed for the purposes of the study. ...
... Due to a myriad of sensors available to researchers through mobile devices, and the rare rate of separation from their owners, they are an invaluable resource of rich user data [10,20]. One such research application for datasets harvested in this manner is the training of intelligent systems for management of mobile noti cations [8]. However, obtaining rich datasets, such as those containing mobile noti cations and contextual user data, from mobile devices does pose a number of problems as there are a number of factors which impede the ease of acquisition. ...
... e di culty being that noti cations are inherently personal and generally contain sensitive information which is unethical to record, use and share. Research in the domain of noti cation management has therefore contained, thus far, many ad-hoc mobile applications which capture a limited number of data points relating to user's incoming noti cations and proceeds to abstract this data to preserve its sensitive nature [8,17,18]. e raw information contained within the noti cations is discarded and intelligent systems are trained on abstracted data which tends to highlight the emphasis currently being placed on ensuring the utmost user privacy and protection. ...
... As an alternative to capturing noti cation data in-the-wild a synthetic noti cation dataset is proposed for use in improving a NMS [8]. In order for the dataset to be useful, it must draw parallels with real-world data. ...
Open-source mobile notification datasets are a rarity in the research community. Due to the sensitive nature of mobile notifications it is difficult to find a dataset which captures their features in such a way that their inherently personal information is kept private. For this reason, the majority of research in the domain of Notification Management requires ad-hoc software to be developed for capturing the data necessary to test hypotheses, train algorithms and evaluate proposed systems. As an alternative, this paper discusses the process, advantages and limitations with harnessing a large-scale mobile usage dataset for deriving a synthetic mobile notification dataset used in testing and improving an intelligent Notification Management System (NMS).
... Mehrotra et al. [9] worked on an in-the-wild dataset that consists of 11,185 notifications from 18 users to develop an intelligent system to recognize useful notifications and to decline if a message will be declined by the user. 3,174 notification data have been used in a study on notification management [11] where notification arrival date, time, notification title, notification message, and application package name has been plotted as features. As stated in [12], contextual data like location, phone events, notification interaction events, etc., have been used to design a notification manager. ...
At present, mobile phone is considered to be a device through which users are always reachable to be communicated wherever and whenever needed via phone calls, SMS, or messaging applications, for instance, WhatsApp, Messenger, etc. Sometimes this accessibility makes people annoyed particularly when someone receives unnecessary and inappropriate notifications in inopportune time (e.g., meeting, seminar, workshop, etc.). Almost every day, we receive bundles of unnecessary SMS from mobile operators and different brands promoting various offers. It distracts us while we are busy with important tasks. Sometimes we miss important messages due to these kinds of promotional messages. This paper presents a prediction model to classify SMS notifications based on users’ preferences. Comparing different machine learning techniques, we have found random forest algorithm gives the highest accuracy (85%).
... are inherently ubiquitous and tend to remain within reach of their owners throughout the day [19], the potential for disruption from mobile push-notifications is much higher than from other notification sources such as desktop computers [15]. For this reason, recent research [1,3,7,8,[11][12][13]18] has been conducted on mobile Notification Management Systems (NMS) which aim to block or delay notifications which are not seen as useful or desired, while still allowing important notifications to be delivered immediately. Currently the majority of these state-of-the-art systems are trained using real user notification data collected in-the-wild and implement some form of supervised learning. ...
... With push notification overload marketers have noticed that "users are more likely to respond to a message with information that directly affects them, as compared to a message that was sent to all of the app's users" 9 . However, this might not be enough and several researchers proposed personalised intelligent notification mechanisms for mobile applications [64], [65] to limit notification disruptions -similarly to what we tested with one application only. ...
Global corporations are characterised by a large number of employees and geographically dispersed offices. Moreover, the competitiveness in the global market requires them to invest in their human resources to be able to remain a step ahead of competition. Implementing large scale classical education in such environments is challenging and costly. Mobile e-learning (m-learning) allows users to tailor their professional training and education to their needs and time constraints. However, in self-paced education it is very hard to keep user retention and engagement. To achieve the latter we have designed and developed an m-learning platform for corporate environments based on the triggering persuasive technology principle that try to incite users in regularly using the platform. We have evaluated the application in-the-wild in corporate environments of differently sized companies with 300 users. Users were subjected to three different conditions: no triggering, simple regular triggering and adaptive triggering. The results show that the use of adaptive triggering in m-learning increases user engagement as well as course completion rates more than simple regular triggering and no triggering.
Image classification is a classic problem in areas pertaining to Computer Vision, Image Processing, and Machine Learning. This paper aims to compare the various Deep Learning Architectures to improve the accuracy of Image Classification to select the best Deep Learning Architecture by implementing and testing various Deep Learning Architectures in combination with Dense Neural Networks. This comparative study helps to improve the accuracy of image separation in both training and testing databases. For targeted training and testing, 3000 training images and 1000 test images were used. The result of the Deep Learning-based classification of images using the platform as Google Colab showed how accurate classification was done by comparing various deep learning architectures.
Notifications on mobile phones alert users about new messages, emails, social network updates, and other events. However, little is understood about the nature and effect of such notifications on the daily lives of mobile users. We report from a one-week, in-situ study involving 15 mobile phones users, where we collected real-world notifications through a smartphone logging application alongside subjective perceptions of those notifications through an online diary. We found that our participants had to deal with 63.5 notifications on average per day, mostly from messengers and email. Whether the phone is in silent mode or not, notifications were typically viewed within minutes. Social pressure in personal communication was amongst the main reasons given. While an increasing number of notifications was associated with an increase in negative emotions, receiving more messages and social network updates also made our participants feel more connected with others. Our findings imply that avoiding interruptions from notifications may be viable for professional communication, while in personal communication, approaches should focus on managing expectations.
Notifications are extremely beneficial to users, but they often demand their attention at inappropriate moments. In this paper we present an in-situ study of mobile interruptibility focusing on the effect of cognitive and physical factors on the response time and the disruption perceived from a notification. Through a mixed method of automated smartphone logging and experience sampling we collected 10372 in-the-wild notifications and 474 questionnaire responses on notification perception from 20 users. We found that the response time and the perceived disruption from a notification can be influenced by its presentation, alert type, sender-recipient relationship as well as the type, completion level and complexity of the task in which the user is engaged. We found that even a notification that contains important or useful content can cause disruption. Finally, we observe the substantial role of the psychological traits of the individuals on the response time and the disruption perceived from a notification.
As the amount of information to users increases with the trends of an increasing numbers of devices, applications, and web services, the new bottleneck in computing is human attention. To minimize users attentional overload, we propose a novel middleware “Attelia” that detects breakpoints of user’s mobile interactions to deliver notifications adaptively. Attelia detects such timings in real-time, using only users phones, without any external sensors, and without any modifications to applications. Our extensive evaluation proved Attelias effectiveness. In-the-wild user study with 30 participants for 16 days showed that, specifically for the users with greater sensitivity for interruptive notification timings, notification scheduling in Attelia’s breakpoint timing reduced users frustration by 28% in users’ real smart phone environments.
Nowadays, notifications are increasingly gaining momentum in our society. New smart devices and appliances are developed everyday with the ability to generate, send and show messages about their status, acquired data and/or information received from other devices and users.
Consequently, the number of notifications received by a user is growing and the tolerance to them could decrease in a short time.
This paper presents a smart notification system that uses machine learning algorithms to adequately manage incoming notifications. According to context awareness and user habits, the system decides: who should receive an incoming notification; what is the best moment to show the notification to the chosen user(s); on which device(s) the chosen user(s) should receive the notification; which is the best way to notify the incoming notification.
After the design of a general architecture, as a first step in building such a system, three different machine learning algorithms were compared in the task of establishing the best device on which the incoming notification should be delivered. The algorithms were applied to a dataset derived from real data provided by the MIT Media Laboratory Reality Mining project, enriched with additional synthetic information.
User models (UMs) allow systems to provide personalized services to their users. Nowadays, UMs are developed ad-hoc, as part of specific applications, thus requiring repetitive development efforts. In this paper, we propose the info-bead user modeling approach, which is based on ideas taken from software engineering in general and component-based software development in particular. The basic standalone unit, the info-bead, represents a single user attribute within time-tagged information-items. An info-bead encapsulates an inference process that uses data received from sensors or other info-beads and yields an information-item value. Having standard interfaces, info-beads can be linked, thus creating info-pendants. Both info-beads and info-pendants can be assembled as needed into complex and abstract user models (UMs) and group models (GMs). The goal of the suggested approach is to ease the modeling process and to allow reuse of info beads developed for one UM in other UMs that need the same information. In order to assess the reusability and collaboration capabilities of the info-bead user modeling approach, we developed a prototype tool that enables UM designers, who are not necessarily software developers, to easily select and integrate info-beads for constructing UMs and GMs. We further demonstrated the use of the approach in a museum environment, for modeling of assistive technology ontology and for user modeling in various specific domains. Finally, we analyzed and assessed the characteristics of the approach with respect to existing generic user modeling criteria.
Modern sensor-equipped smartphones have attracted significant research interest in the pervasive community for recognizing and creating context-aware applications at a personal or community scale level. In this paper, we propose a proof of concept Do-Not-Disturb (DND) service that can a) determine a user's context relevant for DND service from the built-in smartphone sensors and b) correctly predict the DND status based on the given context such as being in a meeting, sleeping, or working at the office. In this preliminary study, we investigate whether sensor data can be clustered to represent user contexts. We use standard machine learning techniques to learn the relationship between a user's context and the corresponding DND status (available or unavailable). Given a user's current context, the DND service predicts a DND status and configures the mobile device accordingly. Our preliminary experiment demonstrates that the proposed system can achieve a prediction accuracy of up to 90% when trained with sufficient data.
Notifications are a core feature of mobile phones. They inform users about a variety of events. Users may take immediate action or ignore them depending on the importance of a notification as well as their current context. The nature of notifications is manifold, applications use them both sparsely and frequently. In this paper we present the first large-scale analysis of mobile notifications with a focus on users’ subjective perceptions. We derive a holistic picture of notifications on mobile phones by collecting close to 200 million notifications from more than 40,000 users. Using a data-driven approach, we break down what users like and dislike about notifications. Our results reveal differences in importance of notifications and how users value notifications from messaging apps as well as notifications that include information about people and events. Based on these results we derive a number of findings about the nature of notifications and guidelines to effectively use them.
Pervasive computing is the vision of technology that is invisibly embedded in our natural surroundings. Users are offered unobtrusive services that require minimal attention. In this paper the Awareness and Notification Service (ANS) is presented that enables to rapidly build applications that inform users about their environment. Additionally, the service offers notifications tailored to the user's preferences and current context. ANS takes a rule based approach based on the event-condition-action pattern. Users specify when and what should be notified to them by using a convenient ANS rule language. This flexible mechanism allows to rapidly develop applications that provide context-aware notifications without the need to write programming code to activate rules, nor to implement personalized notifications
Desktop notifications are designed to provide awareness of information while a user is attending to a primary task. Unfortunately the awareness can come with the price of disruption to the focal task. We review results of a field study on the use and perceived value of email notifications in the workplace. We recorded users' interactions with software applications for two weeks and studied how notifications or their forced absence influenced users' quest for awareness of new email arrival, as well as the impact of notifications on their overall task focus. Results showed that users view notifications as a mechanism to provide passive awareness rather than a trigger to switch tasks. Turing off notifications cause some users to self interrupt more to explicitly monitor email arrival, while others appear to be able to better focus on their tasks. Users acknowledge notifications as disruptive, yet opt for them because of their perceived value in providing awareness.
In the molecular era the management of cancer is no more a plan based on simple guidelines. Clinical findings, tumor characteristics, and molecular markers are integrated to identify different risk categories, based on which treatment is planned for each individual case. This paper aims at developing a fuzzy decision support system (DSS) to guide the doctors for the risk stratification of breast cancer, which is expected to have a great impact on treatment decision and to minimize individual variations in selecting the optimal treatment for a particular case. The developed system was based on clinical practice of Oncology Center Mansoura University (OCMU). This system has six input variables (Her2, hormone receptors, age, tumor grade, tumor size, and lymph node) and one output variable (risk status). The output variable is a value from 1 to 4; representing low risk status, intermediate risk status and high risk status. This system uses Mamdani inference method and simulation applied in MATLAB R2009b fuzzy logic toolbox.
The medical consultation system MedFrame/CADIAG-IV is a successor of the prior CADIAG projects. It is the result of a complete redesign to account for today's demands on state-of-the-art software. Its knowledge representation and inference process are based on fuzzy set theory and fuzzy logic. Fuzzy sets are used for conversions from measured numeric values and observational data into symbolic ones. Medical relationships between findings, diseases, and therapies, the rules, are represented by fuzzy relations, that express positive or negative associations. Findings, diseases, and therapies are organised in hierarchies.
this article, we describe principles that can help organizations develop usable data quality metrics
Mobile computing poses a big challenge to user modeling -- how to provide users with personalized services anywhere and anytime, without requiring the user to bootstrap a user model (UM) from scratch every time. The basic, abstract challenge is to maintain an up to date UM. Today's technology provides solutions to this abstract challenge -- the UM may be stored in the cloud or on the user's mobile device (or even both). However, the devil is in the details -- how can the environment or service provider access the UM and get the needed information from the model? How can the UM (or the user-modeling component) respond to continuous requests for personal data required by services in ubiquitous computing? This paper presents the idea and work in progress of "Mobile Info-Beads User Model" -- an extension of the "Info-Beads User Modeling approach" to the mobile world, where user-modeling components are exchanged and composed for providing the current service with the personal data needed for personalization. When such approach is adopted, it paves the way for dynamic construction of UMs, as needed, when needed, hence providing the infrastructure needed for adapting and extending UMs "on the fly" for personalized ubiquitous computing applications.
We present an observational study of how notifications are handled by collocated groups, in the context of a collaborative mobile photo-taking exercise. Interaction analysis of video recordings is used to uncover the methodical ways in which participants manage notifications, establishing and sustaining co-oriented interaction to coordinate action, such as sharing notification contents and deciding on courses of action. Findings highlight how embodied and technological resources are collectively drawn upon in situationally nuanced ways to achieve the management of notifications delivered to cohorts. The insights can be used to develop an understanding of how interruptions are dealt with in other settings, and to reflect on how to support notification management within collocated groups by design.
Following the increase in of the information available on the Web, the diversity of its users and the complexity of Web applications, researchers started developing adaptive Web systems that tailored their appearance and behavior to each individual user or user group. Adaptive systems were designed for different usage contexts, exploring different kinds of personalization. Web personalization has evolved into a large research field attracting scientists from different communities such as hypertext, user modeling, machine learning, natural language generation, information retrieval, intelligent tutoring systems, cognitive science, and Web-based education.
This state-of-the-art survey provides a systematic overview of the ideas and techniques of the adaptive Web and serves as a central source of information for researchers, practitioners, and students. The volume constitutes a comprehensive and carefully planned collection of chapters, mapping out the most important areas of the adaptive Web, each solicited from experts and leaders in the field.
The largest part of the book focuses on personalization techniques, namely the modeling side of personalization (Chaps. 1-5), and on adaptation, (Chaps. 6-14). The technique-focused part is complemented by four domain-oriented chapters in the third section of the book (Chaps. 15-18). The last section is devoted to recently emerging topics; it provides a prospective view of the new ideas and techniques that are moving rapidly into the focus of the adaptive Web community and gives the reader a glimpse into the not-so-distant future.
Management of uncertainty is an intrinsically important issue in the design of expert systems because much of the information in the knowledge base of a typical expert system is imprecise, incomplete or not totally reliable.In the existing expert systems, uncertainty is dealt with through a combination of predicate logic and probability-based methods. A serious shortcoming of these methods is that they are not capable of coming to grips with the pervasive fuzziness of information in the knowledge base, and, as a result, are mostly ad hoc in nature. An alternative approach to the management of uncertainty which is suggested in this paper is based on the use of fuzzy logic, which is the logic underlying approximate or, equivalently, fuzzy reasoning. A feature of fuzzy logic which is of particular importance to the management of uncertainty in expert systems is that it provides a systematic framework for dealing with fuzzy quantifiers, e.g., most, many, few, not very many, almost all, infrequently, about 0.8, etc. In this way, fuzzy logic subsumes both predicate logic and probability theory, and makes it possible to deal with different types of uncertainty within a single conceptual framework.In fuzzy logic, the deduction of a conclusion from a set of premises is reduced, in general, to the solution of a nonlinear program through the application of projection and extension principles. This approach to deduction leads to various basic syllogisms which may be used as rules of combination of evidence in expert systems. Among syllogisms of this type which are discussed in this paper are the intersection/product syllogism, the generalized modus ponens, the consequent conjunction syllogism, and the major-premise reversibility rule.
CADIAG-2 is a successful expert system assisting in the differential diagnosis in internal medicine. With its aid, conjectures about present diseases are derived from possibly vague information about a patient's symptoms. In this paper we provide a mathematical formalisation of the inferential mechanism of CADIAG-2. A Gentzen-style calculus for the resulting logic is introduced and used to compare the system's behaviour with t-norm based fuzzy logics.
Fuzzy set theory and fuzzy logic are a highly suitable and applicable basis for developing knowledge-based systems in medicine for tasks such as the interpretation of sets of medical findings, syndrome differentiation in Eastern medicine, diagnosis of diseases in Western medicine, mixed diagnosis of integrated Western and Eastern medicine, the optimal selection of medical treatments integrating Western and Eastern medicine, and for real-time monitoring of patient data. This was verified by trials with the following systems that were developed by our group in Vietnam: a fuzzy Expert System for Syndromes Differentiation in Oriental Traditional Medicine, an Expert System for Lung Diseases using fuzzy logic, Case Based Reasoning for Medical Diagnosis using fuzzy set theory, a diagnostic system combining disease diagnosis of Western Medicine with syndrome differentiation of Oriental Traditional Medicine, a fuzzy system for classification of Western and Eastern medicaments and finally, a fuzzy system for diagnosis and treatment of integrated Western and Eastern Medicine.
In this paper we propose to use context information obtained from body-worn sensors to mediate notifications for a wearable computer In particular we introduce a model which uses two axes, namely personal and social interruptability of the user in order to decide both whether or not to notify the user and to decide which notification modality to use. Rather than to model and recognize the complete context of the user we argue that personal and social interruptability can be derived directly from various sensors by the combination of tendencies. First experimental results show the feasibility of the approach using acceleration, audio, and location sensors.
- L L Pipino
- Y W Lee
- R Y Wang
L. L. Pipino, Y. W. Lee and R. Y. Wang, " Data Quality Assessment, "
Commun. ACM, vol. 45, no. 4, pp. 211-218, Apr 2002.
A context and user aware smart notification system
- F Corno
- L D Russis
- T Montanaro
F. Corno, L. D. Russis and T. Montanaro, "A context and user aware
smart notification system," in Internet of Things (WF-IoT), 2015 IEEE
2nd World Forum on, 2015.
- E Dim
- T Kuflik
- I Reinhartz-Berger
E. Dim, T. Kuflik and I. Reinhartz-Berger, "When User Modeling
Intersects Software Engineering: The Info-bead User Modeling
Approach," User Modeling and User-Adapted Interaction, vol. 25, no. 3,
pp. 189-229, Aug 2015.
- B Sageder
- K Boegl
- K.-P Adlassnig
- G Kolousek
- B Trummer
B. Sageder, K. Boegl, K.-P. Adlassnig, G. Kolousek and B. Trummer,
"The knowledge model of MedFrame/CADIAG-IV.," Studies in health
technology and informatics, vol. 43, pp. 629-633, 1996.