Questioning the Question -- Addressing the Answerability of Questions in Community Question-Answering

Shah, Chirag; Kitzie, Vanessa; Choi, Erik

doi:10.1109/hicss.2014.180

Cited by 20 publications

(14 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Several approaches exist to identifying Best Answers. Liu et al () generalized these approaches into an Asker Satisfaction Prediction (ASP) framework, including textual and semantic features of questions and answers, history of answer satisfaction by category, and past activity history of askers and answerers (Belkin et al, ; Jeon et al, ; Liu et al, ; Agichtein et al, ; Shah & Pomerantz, ; Shah, Kitzie, & Choi, ). Other studies have attempted to both add classification features and use other evaluative baselines for answer quality using human‐based assessments.…”

Section: Introductionsupporting

confidence: 83%

Building a parsimonious model for identifying best answers using interaction history in community Q&A

Shah

2015

Proc. Assoc. Info. Sci. Tech.

Self Cite

View full text Add to dashboard Cite

Evaluating answer quality or identifying/predicting which answer would be selected as the best for a given question is an important problem in community-based Q&A services. In this article we introduce new interaction-based features depicting the amount of distinct interactions between an asker and answerer over time, in order to predict whether an answer will be selected as Best Answer or not within Yahoo! Answers. Through a series of experiments ran on a data set of 23,218 question-answer pairs, we determined that after the data was first run using a model trained on textual features, and then the failed cases re-run with a model trained on interaction features, we were able to significantly improve the performance of the original model in identifying these difficult cases. In addition, when compared to models using often five to seven times the amount of features and requiring a large amount of computational effort, our model performed at to above the same evaluative measures. This suggests that future classification models can be made more parsimonious and handle larger datasets using less computational effort by developing a two-step classifier that includes interaction history as a feature.

show abstract

Section: Introductionsupporting

confidence: 83%

Building a parsimonious model for identifying best answers using interaction history in community Q&A

Shah

2015

Proc. Assoc. Info. Sci. Tech.

Self Cite

View full text Add to dashboard Cite

show abstract

“…A core issue in the literature on social Q&A interactions and online information seeking is the idea that some information requests are better than others [60][61][62][63] with good questions seen as those which are more likely to receive an answer [60]. Choi et al [60], looking at factual questions on Yahoo Answers!, found that textual features, such as the level of clarity in a question, can be important in predicting if a question will receive an answer of not.…”

Section: Good and Bad Questions: Answer Success And Failurementioning

confidence: 99%

“…[66]. Shah et al proposed different types of questions on social Q&A sites: factual, advice, opinion seeking and social questions and demonstrated that adding information on question type could improve the performance of automatic classifiers based on textual features [63]. The implications from much of this work are that systems may wish to understand what is the real need faced by the poster and tackle that.…”

Section: Moderation and Automatic Classificationmentioning

confidence: 99%

The language of information need: Differentiating conscious and formalized information needs

Ruthven¹

2019

Information Processing & Management

View full text Add to dashboard Cite

Information need is a fundamental concept within Information Science. Robert Taylor's seminal contribution in 1968 was to propose a division of information needs into four levels: the visceral, conscious, formalized and compromised levels of information need. Taylor's contribution has provided much inspiration to Information Science research but this has largely remained at the discursive and conceptual level. In this paper, we present a novel empirical investigation of Taylor's information need classification. We analyse the linguistic differences between conscious and formalized needs using several hundred postings to four major Internet discussion groups. We show that descriptions of conscious needs are more emotional in tone, involve more sensory perception and contain different temporal dimensions than descriptions of formalized needs. We show that it is possible to differentiate levels of information need based on linguistic patterns and that the language used to express information needs can reflect an individual's understanding of their information problem. This has implications for the theory of information needs and practical implications for supporting moderators of online news groups in responding to information needs and for developing automated support for classifying information needs.

show abstract

“…The ratio of words of four or more letters to the total number of words [40,43] The ratio of words of six or more letters to the total number of words…”

Section: Richnessmentioning

confidence: 99%

Defining High-Quality Answers on a Chinese Tourism Q&A Platform in Terms of Information Needs

Song

Liu

et al. 2021

Sustainability

View full text Add to dashboard Cite

Travel Question and Answer (Q&A) platforms are widely used by travelers to exchange information online. This paper examined one such platform, Ctrip Q&A, to better understand the features of high-quality answers. The questions were first categorized into four types of information required by tourists (knowledge, efficiency, risk, and hedonic types) and then subdivided into more concrete elements, such as accommodation, food, and climate. The characteristics of high-quality answers on the platform were identified and a regression analysis was conducted to determine their importance. The study found that negative emotional expressions negatively impacted the quality of answers while the usefulness, richness, and amount of information had a positive effect. The usefulness of the answer content and quantity of information consistently predicted high-quality answers, whereas richness and conciseness were inconsistently related. Finally, the characteristics of high-quality answers to questions about different elements of tourism in the same category of tourist information varied significantly. This study may help travelers select and provide high-quality answers on tourism Q&A platforms. It may also assist platform operators in optimizing the features of their platforms, thereby improving user satisfaction and overall quality.

show abstract

Questioning the Question -- Addressing the Answerability of Questions in Community Question-Answering

Cited by 20 publications

References 24 publications

Building a parsimonious model for identifying best answers using interaction history in community Q&A

Building a parsimonious model for identifying best answers using interaction history in community Q&A

The language of information need: Differentiating conscious and formalized information needs

Defining High-Quality Answers on a Chinese Tourism Q&A Platform in Terms of Information Needs

Contact Info

Product

Resources

About