Figure 1 - uploaded by Daphne R. Raban
Content may be subject to copyright.
A question in Yahoo! Answers.  

A question in Yahoo! Answers.  

Source publication
Conference Paper
Full-text available
Question and answer (Q&A) sites such as Yahoo! Answers are places where users ask questions and others answer them. In this paper, we investigate predictors of answer quality through a comparative, controlled field study of responses provided across several online Q&A sites. Along with several quantitative results concerning the effects of factors...

Context in source publication

Context 1
... Answers presents an example of such an interface (see ). Figure 1 Three Types of Q&A Sites ...

Similar publications

Chapter
Full-text available
Systematic experience reuse supports humans to make better use of their own experience, and more importantly, to make use of experience which is made by a community rather than which is made by an individual. Experience management does not come for free. It involves changing existing or introducing new activities at well as IT technology that effic...

Citations

... Therefore, in order to stock appropriate answer statements, it would be vital to demand respondents to be able to provide appropriate answers. For the purpose of solving the issues described earlier, there have been abundant prior works researching Q&A sites [2][3][4][5][6][7][8][9] using textual features or link analysis. However, these works have not taken the tendencies of the written styles of users into consideration. ...
... For some of the dataset, their proposal model showed good performance in recommending new questions to experts. Harper et al. sought for predictors of answer quality through a comparative and controlled field study of responses provided across several online Q&A sites [3]. The comparison analysis has shown that each Q&A site had individual tendencies and features about user contributions to the sites. ...
... Meanwhile, this work focuses on using impressions in addition to textual features. Moreover, in spite of several previous studies that introduce users to answer statements as Summarization of the Methodology of Applying N-gram to Obtain Factor Scores of Q&A Statements summarized [2][3][4][5][6][7][8][9], a methodology to introduce appropriate respondents to a questioner has yet to be established. Hence, by using the impression of statements, the purpose of this work is to introduce appropriate respondents to a questioner. ...
... Users could pay between $2 and $200 for a question and could also tip answerers. Studies by [46] showed that paying more for an answer in GA produced higher-quality answers, longer answers, and higher effort from answerers. However, a field study by [47] contradicted this view, finding that paying more only resulted in longer answers, not better quality answers. ...
Preprint
Full-text available
Q&A websites rely on user-generated responses, with incentives such as reputation scores or monetary rewards often offered. While some users may find it intrinsically rewarding to assist others, studies indicate that payment can improve the quality and speed of answers. However, traditional payment processors impose minimum thresholds that many Q&A inquiries fall below. The introduction of Bitcoin enabled direct digital value transfer, yet frequent micropayments remain challenging. Recent advancements like the Lightning Network now allow frictionless micropayments by reducing costs and minimising reliance on intermediaries. This development fosters an "Internet of Value," where transferring even small amounts of money is as simple as sharing data. This study investigates integrating Lightning Network-based micropayment strategies into Q&A platforms, aiming to create a knowledge market free of minimum payment barriers. A survey was conducted to address the gap below the $2 payment level identified in prior research. Responses confirmed that incentives for asking and answering weaken as payments decrease. Findings reveal even minimal payments, such as {\pounds}0.01, significantly encourage higher quality and effort in responses. The study recommends micropayment incentives for service-oriented applications, particularly Q&A platforms. By leveraging the Lightning Network to remove barriers, a more open marketplace can emerge, improving engagement and outcomes. Further research is needed to confirm if users follow through on reported intentions when spending funds.
... Though web traffic and question quality are closely tied, past studies mainly focused on evaluating answers over questions (Harper et al., 2008;Roy et al., 2018;Wang et al., 2013). A single prior work explored evaluative attributes of CQA questions (Sha et al., 2020), yet the specific impacts of question quality on website performance remain unclear, highlighting the need for further research in this area.in ...
... Furthermore, compared to the summarization task, LLM judges exhibit a greater degree of length bias on the multi-turn conversation task (HH-RLHF-Helpfulness dataset). Generally, longer responses tend to provide more detailed and comprehensive answers, which are more favored by humans compared to shorter ones [51,52]. We suspect that the length bias results from the over-alignment of commercial models with human preferences. ...
Preprint
Full-text available
Alignment approaches such as RLHF and DPO are actively investigated to align large language models (LLMs) with human preferences. Commercial large language models (LLMs) like GPT-4 have been recently employed to evaluate and compare different LLM alignment approaches. These models act as surrogates for human evaluators due to their promising abilities to approximate human preferences with remarkably faster feedback and lower costs. This methodology is referred to as LLM-as-a-judge. However, concerns regarding its reliability have emerged, attributed to LLM judges' biases and inconsistent decision-making. Previous research has sought to develop robust evaluation frameworks for assessing the reliability of LLM judges and their alignment with human preferences. However, the employed evaluation metrics often lack adequate explainability and fail to address the internal inconsistency of LLMs. Additionally, existing studies inadequately explore the impact of various prompt templates when applying LLM-as-a-judge methods, which leads to potentially inconsistent comparisons between different alignment algorithms. In this work, we systematically evaluate LLM judges on alignment tasks (e.g. summarization) by defining evaluation metrics with improved theoretical interpretability and disentangling reliability metrics with LLM internal inconsistency. We develop a framework to evaluate, compare, and visualize the reliability and alignment of LLM judges to provide informative observations that help choose LLM judges for alignment tasks. Our results indicate a significant impact of prompt templates on LLM judge performance, as well as a mediocre alignment level between the tested LLM judges and human evaluators.
... This discrepancy is a topic for future research. However, the results of this research are consistent with a study on expression that found no significant correlation between rhetorical strategies and predicting answer outcomes (Harper, et al., 2008). ...
Article
Full-text available
This study proposes an exploratory methodology for analyzing participation in Electronic Networks of Practice, focusing on a network of 500 specialists in breast cancer treatment as a case study. The research addresses a critical gap in understanding how conversational dynamics influence participation in professional forums. Conversation measures and participation measures reported in the literature are identified and categorized, and new ones are suggested. This study by correlation analysis reveals three key participation triggers: initiator features, timing of initiation, and feedback, while also identifying non-associated variables like conversation topic and expression. Clustering categorizes conversations into four types, each exhibiting participation patterns and attributes. Social network analysis further reveals the network’s structure and participants’ interrelations. These findings provide a comprehensive understanding of participation in Electronic Networks of Practice and show the methodology’s broader applicability in analyzing professional networks. The study’s insights are vital for practitioners and theorists by offering a perspective on fostering knowledge sharing.
... Some studies have examined replier characteristics, including reputation score , expertise (Edwards, 2008), and experience (Shah & Pomerantz, 2010). Additionally, question-level factors such as question length (Shah & Pomerantz, 2010), attribute category (Harper et al., 2008), and reward size (Jeon et al., 2010) have also been examined to determine how they affect the perceived helpfulness of answers. Subsequently, a greater portion of the literature has concentrated on the characteristics of answer content. ...
... First, we account for two variables that reflect the detailed degree of question content Zhang et al., 2020): SentenceSum j , representing the total number of sentences in the body of question j , and QueContentWord j , representing the total number of words in the content of question j (normalized) (Peng et al., 2020). Second, to account for readers' tendency to favor answers under paid questions, we also consider whether a question had been offered a financial reward (Harper et al., 2008;Jeon et al., 2010). We define Reward j , which is a dummy variable, as equal to 1 if question j offers a financial reward (and 0 otherwise). ...
... We define Reward j , which is a dummy variable, as equal to 1 if question j offers a financial reward (and 0 otherwise). (3) Finally, considering the impact of question breadth (Harper et al., 2008;Jeon et al., 2010), we control for the total number of subtags attached to a question, which is defined as TagSum j , denoting the total number of subtags attached to question j . ...
Article
Full-text available
Given the unwieldy glut of information in online health question-answer (Q&A) service, it is essential to understand what constitutes helpful answers in the medical domain. Despite the fact that studies have examined the impacts of answer content factors on answer helpfulness, there are two gaps that need further analysis. First, the empirical results of the existing relevant studies on the effect of answer emotion are inconsistent. Second, prior studies only have examined the independent impacts of answer content factors and question content cues on answer helpfulness. To fill this gap, a research model reflecting the impacts of emotional content and question-answer congruence on answer helpfulness was developed and empirically examined. Our empirical analyses confirm that emotional content and answer helpfulness are related to one another in the form of an inverted U-shape and indicate that two types of question-answer congruence (emotional intensity congruence and linguistic style matching) positively affect answer helpfulness. Theoretical and practical implications are discussed.
... Extensive studies have been conducted, employing discriminative features and classifiers to identify the best answers. For example, earlier works have indicated that answer length prevails as a dominant feature among others in identifying the best answer [78]. The hallmark of a high-quality answer encompasses traits like clarity, visual adequacy, and the provision of necessary references and citations to external resources. ...
Article
Full-text available
The growth of digital platforms has led to the proliferation of Online Communities, providing individuals with opportunities to seek help and share knowledge. A key challenge of help-related platforms that address technical questions (i.e., utilitarian, rather than opinion or supportive) is to ensure the contributions address seekers’ specific information needs. Despite growing academic interest in such platforms, research has mainly focused on factors that influence the quantity of contributions, ignoring whether these contributions effectively helped the seekers. To fill this research gap, this study draws upon theories of self-determination and motivation crowding to examine contributing behaviors that result in successful helping. By analyzing a rich dataset collected from an online Q&A platform, we find that gains in a help provider’s past rewards positively influence the success of contribution. Further, while previous studies suggest that external rewards result in a high quantity of contribution, our findings show that an inflated frequency of contribution leads to a crowding-out effect. Specifically, the contribution frequency has a curvilinear relationship with the success of the contribution. Taken together, these findings demonstrate there is a need to revisit the gamification mechanism on help-related platforms to ensure the success of knowledge contribution. This is crucial for the sustainability of these platforms as low-quality answers can lead users to mistrust and eventually leave the platform.
... chance of a response [30], [31]. Many studies have examined the answerability of questions on Q&A sites [29], [42], [44]- [48]. For example, Harper et al. [44] explored the variables that affect answer outcomes (such as number, length, effort, and quality of answers) and found that question topics, question types, levels of reward, and the site itself significantly affected one or more of these outcomes. ...
... Many studies have examined the answerability of questions on Q&A sites [29], [42], [44]- [48]. For example, Harper et al. [44] explored the variables that affect answer outcomes (such as number, length, effort, and quality of answers) and found that question topics, question types, levels of reward, and the site itself significantly affected one or more of these outcomes. Another study showed that the topics, uniqueness, and urgency of questions significantly affected the possibility of receiving answers [45]. ...
Article
Although security and privacy technologies are incorporated into every device and service, the complexity of these concepts confuses non-expert users. Prior research has shown that non-expert users ask strangers for advice about digital media use online. In this study, to clarify the security and privacy concerns of non-expert users in their daily lives, we investigated security- and privacy-related question posts on a Question-and-Answer (Q&A) site for non-expert users. We conducted a thematic analysis of 445 question posts. We identified seven themes among the questions and found that users asked about cyberattacks the most, followed by authentication and security software. We also found that there was a strong demand for answers, especially for questions related to privacy abuse and account/device management. Our findings provide key insights into what non-experts are struggling with when it comes to privacy and security and will help service providers and researchers make improvements to address these concerns.
... For instance, it may be difficult for automatic moderation systems to reliably identify minor manifestations of toxic behaviour and prevent false positives [43]. Strict moderating guidelines may reduce the variety of viewpoints and deter members of underrepresented groups from participating [44]. ...
Article
Full-text available
A renowned online community known as Quora enables members to post queries, receive insightful responses, and share knowledge. The capacity of Quora to find related questions based on a user's search is a distinctive feature that makes it simple for users to access pertinent information and add it to the platform's knowledge base. The retrieval of comparable questions from Quora is the topic of this paper. We assess various systems that classify related queries and quickly deliver pertinent responses to information searchers. Our assessment of machine learning and natural language processing methods focuses on how well these methods work when obtaining queries from the large Quora question database that serves related objectives. Our thorough research paper provides a summary of the literature on comparable question retrieval in Quora while highlighting the benefits and drawbacks of various approaches. Our evaluation identifies prospective topics for more research and development and acts as a guide for future scholars interested in this field. By enhancing similar question retrieval on Quora, we hope to encourage knowledge-sharing and community development on this important platform. Users can find the most pertinent responses to their inquiries on Quora by using the study's findings.
... A significant association was found between question type and answerers' effort (Harper et al., 2008). Answering informational questions is providing others with advice or facts. ...
Article
Social question-and-answer (Q&A) communities are becoming increasingly important for knowledge acquisition. However, some users withhold knowledge , which can hinder the effectiveness of these platforms. Based on social exchange theory, the study investigates how different types of questions influence knowledge withholding, with question difficulty and user anonymity as boundary conditions. Two experiments were conducted to test hypotheses. Results indicate that informational questions are more likely to lead to knowledge withholding than conversational ones, as they elicit more fear of negative evaluation and fear of exploitation. The study also examines the interplay of question difficulty and user anonymity with question type. Overall, this study significantly extends the existing literature on counterproductive knowledge behavior by exploring the antecedents of knowledge withholding in social Q&A communities.