Exploiting user feedback to learn to rank answers in q&amp;a forums

Dalip, Daniel Hasan; Gonçalves, Marcos André; Cristo, Marco; Calado, Pável

doi:10.1145/2484028.2484072

Cited by 78 publications

(52 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Ratio of all likes received by the author coming from his friends and followers [10][11][12][13] Total number of tips posted by the author's social network a [14][15][16][17] Number of likes given by author's social network (in any tip) a [18][19][20][21] Fraction of all likes received by the tip's author that are associated with tips posted at the same venue of the current tip but after it was posted a 22…”

Section: Prediction Results Varying Target Prediction Window δ 505mentioning

confidence: 99%

“…10 Accurate tip popularity predictions can drive the design of automatic tip filtering and recommendation schemes, which in turn 11 can help users find tips that are potentially more valuable more easily. Business owners may also benefit from such predictions 12 as they are able to more quickly identify (and fix) aspects of their services or products that may affect revenues most. 13 However, as in longer review systems, the number of tips on a single product or service may be large and vary greatly in 14 quality [22,26], which makes it hard for users to find helpful reviews.…”

Section: Introductionmentioning

confidence: 99%

“…2(a) and (b). This is because the precision for the smaller 371 We now analyze the relative importance of each feature for prediction accuracy by using the Information Gain (IG) feature 372 selection technique [12] to rank features according to their discriminative capacity for the prediction task. We here use the OLS 9th positions of the ranking, respectively.…”

mentioning

confidence: 99%

“…689 We also emphasize that features 2, 7, 9, 10-24, 30-31 have not been explored before, being thus new contributions of this 690 work. Other features, notably features 32, 33, 38, 41, 55, 60, 65, 67, 69-71, 73-79, 81, 90-125 have been adapted from similar 691 concepts in other domains (e.g., question-answering sites, Wikipedia, and YouTube) [1,11,12,30,36]. 692 …”

mentioning

confidence: 99%

See 3 more Smart Citations

Predicting the popularity of micro-reviews: A Foursquare case study

Vasconcelos

Almeida

Gonçalves

2015

Information Sciences

Self Cite

View full text Add to dashboard Cite

Section: Prediction Results Varying Target Prediction Window δ 505mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

Predicting the popularity of micro-reviews: A Foursquare case study

Vasconcelos

Almeida

Gonçalves

2015

Information Sciences

Self Cite

View full text Add to dashboard Cite

“…[10] H. Dalip, M. A. Gonc¸alves, M. Cristo, and P. Calado (2013) has represented Exploiting user feedback to learn to rank answers in qa forums: A case study with stack overflow [10].The author proposes a learning to rank (L2R) approach for ranking answers in Q&A forums. In particular, we adopt an approach based on Random Forests and represent query and answer pairs using eight different groups of features.…”

mentioning

confidence: 99%

Survey Paper on a Pairwise Learning to Rank Model for Answer Selection in Community Question Answer (CQA) System

Patil¹,

Shah²,

Ukhale³

et al. 2018

IJCA

View full text Add to dashboard Cite

To find the similar questions is very difficult in Question Answering (QA) System. Because each question in the returned candidate pool consists of multiple answers, and hence users get trouble to browse a lot before finding the correct one. To overcome this problem, we construct a novel approach a novel Pair wise learning to rank model i.e PLANE which can quantitatively rank answer candidates from the relevant question pool. Specifically, it comprises two components i.e. one offline learning component and one online search component. In the offline learning component, we first consequently set up the positive, neutral, and negative training samples in the forms of preference pairs guided by our datadriven observations. We at that point display a novel model to together consolidate these three sorts of preparing tests and the closed-form solution of this model is determined. In the online search component, we initially gather a pool of answer candidates for the given question by means of discovering its comparable or similar questions. We at that point sort the appropriate answer candidates by utilizing the offline trained model to judge the preference orders. We also design recommendation system, in which best solution is recommended. The system also provides facilities like bookmarking as well as sends best answer on email. Our model is robust as well as achieves better performance than several state-of-the-art answer selection baselines.

show abstract

A general multiview framework for assessing the quality of collaboratively created content on web 2.0

Dalip

Gonçalves

Cristo

et al. 2016

Asso for Info Science & Tech

Self Cite

View full text Add to dashboard Cite

User-generated content is one of the most interesting phenomena of current published media, as users are now able not only to consume, but also to produce content in a much faster and easier manner. However, such freedom also carries concerns about content quality. In this work, we propose an automatic framework to assess the quality of collaboratively generated content. Quality is addressed as a multidimensional concept, modeled as a combination of independent assessments, each regarding different quality dimensions. Accordingly, we adopt a machine-learning (ML)based multiview approach to assess content quality. We perform a thorough analysis of our framework on two different domains: Questions and Answer Forums and Collaborative Encyclopedias. This allowed us to better understand when and how the proposed multiview approach is able to provide accurate quality assessments. Our main contributions are: (a) a general ML multiview framework that takes advantage of different views of quality indicators; (b) the improvement (up to 30%) in quality assessment over the best state-of-the-art baseline methods; (c) a thorough feature and view analysis regarding impact, informativeness, and correlation, based on two distinct domains.

show abstract

Exploiting user feedback to learn to rank answers in q&a forums

Cited by 78 publications

References 20 publications

Predicting the popularity of micro-reviews: A Foursquare case study

Predicting the popularity of micro-reviews: A Foursquare case study

Survey Paper on a Pairwise Learning to Rank Model for Answer Selection in Community Question Answer (CQA) System

A general multiview framework for assessing the quality of collaboratively created content on web 2.0

Contact Info

Product

Resources

About