Exploiting social context for review quality prediction

Lu, Yue; Tsaparas, Panayiotis; Ntoulas, Alexandros; Polanyi, Livia

doi:10.1145/1772690.1772761

Cited by 238 publications

(120 citation statements)

References 14 publications

Supporting

Mentioning

119

Contrasting

Unclassified

Order By: Relevance

“…They found that tense and time expressions are the most useful linguistic clues to assess the customer's previous purchase experiences. This approach is similar to the work done by the group in [18], however there is a difference, because this work uses the reviewer's social context information based on the social network-based review website, such as the PageRank score of the author. Features used in this work improve system performance, however, they have limitations similar to the "helpfulness votes".…”

Section: Classification and Other Methodsmentioning

confidence: 99%

“…Lu, Tsaparas, Ntoulas, and Polanyi [18], investigate if the social context of reviews can enhance the performance of quality prediction. In their view, important information can be obtained from the social context about the quality of reviewers, which affects the quality of their reviews.…”

Section: Quality As Regression Problemmentioning

confidence: 99%

“…In order to incorporate social context in predicting review quality, regularization constraints, based on a set of experimentallyvalidated hypotheses, were employed. An example is the "author constraint hypothesis", which assumes that reviews from the same author are similar in quality [18]. The results show accuracy improvement in predicting the quality of reviews using a text-based classifier (linear regression model).…”

Section: Quality As Regression Problemmentioning

confidence: 99%

See 2 more Smart Citations

A Survey of Quality Prediction of Product Reviews

Almagrabi¹,

Malibari²,

McNaught³

2015

ijacsa

View full text Add to dashboard Cite

Abstract-With the help of Web-2.0, the Internet offers a vast amount of reviews on many topics and in different domains. This has led to an explosive growth of product reviews and customer feedback, which presents the problem of how to handle the abundant volume of data. It is an expensive and time-consuming task to analyze this huge content of opinions. Therefore, the need for automated sentiment analysis systems is vital. However, these systems encounter many challenges; assessing the content quality of the posted opinions is an important area of study that is related to sentiment analysis. Currently, review helpfulness is assessed manually; however the task of automatically assessing it has gained more attention in recent years. This paper provides a survey of approaches to the challenge of identifying the content quality of product reviews.

show abstract

Section: Classification and Other Methodsmentioning

confidence: 99%

Section: Quality As Regression Problemmentioning

confidence: 99%

Section: Quality As Regression Problemmentioning

confidence: 99%

See 1 more Smart Citation

A Survey of Quality Prediction of Product Reviews

Almagrabi¹,

Malibari²,

McNaught³

2015

ijacsa

View full text Add to dashboard Cite

show abstract

“…User-user and post-post correlations in social media are widely used to improve various tasks such as feature selection [73], sentiment analysis [38,45] and recommendation [49]. Next, I propose a novel way to compute the user-user and post-post correlations to tackle the sparsity problem further using users' actions on public posts…”

Section: Modeling Correlationsmentioning

confidence: 99%

“…I define Ψ(i, j) to measure the user-user correlation coefficients between u i and u j . There are many ways to measure user-user correlation, such as similarity of users' behavior [50] and connections in social networks [49]. In this chapter, I choose the similarity of users' historical behavior to measure user-user correlations.…”

Section: Modeling User-user Correlationsmentioning

confidence: 99%

Managing a User's Vulnerability on a Social Networking Site

Gundecha¹

2015

View full text Add to dashboard Cite

Perceived value of information sharing in online environments: User engagement and social reputation

Choi

Shah

2014

Proc of Assoc for Info

View full text Add to dashboard Cite

Consumer reviews have received much attention both in academia and industry, since it has been found that they have significant effects on people's decision‐making behaviors relating to online shopping or choosing services. The readers/visitors on various online communities where such reviews are posted evaluate the helpfulness of a review based on the extent to which the review is helpful in their information seeking and decision‐making tasks. However, it is often not clear what makes a review useful to a reader. Several previous works have analyzed textual features relating to reviews to determine their usefulness, but given that most of these reviews are generated and shared through community‐based or social media sites, the features that incorporate social features and social engagement could have an important role to play in making a review useful. The work described in this paper uses data available from Yelp, a social review and recommendation site, to build models that examine the impact of various features, including basic, reviewer's engagement, social reputation, and business type, based on the number of votes for helpfulness a review receives. Our findings suggest that reviewer's social contextual information (engagement and reputation) is influential in affecting perceived helpfulness of a review and improves the accuracy of prediction model. The findings also suggest that business type (restaurants vs. transportations) affects the ways in which consumers consider a review.

show abstract

Exploiting social context for review quality prediction

Cited by 238 publications

References 14 publications

A Survey of Quality Prediction of Product Reviews

A Survey of Quality Prediction of Product Reviews

Managing a User's Vulnerability on a Social Networking Site

Perceived value of information sharing in online environments: User engagement and social reputation

Contact Info

Product

Resources

About