Augmenting Chinese Online Video Recommendations by Using Virtual Ratings Predicted by Review Sentiment Classification

Abstract

Abstract-In this paper we aim to resolve the recommendation problem by using the virtual ratings in online environments when user rating information is not available. As a matter of fact, in most of current websites especially the Chinese video-sharing ones, the traditional pure rating based collaborative filtering recommender methods are not fully qualified due to the sparsity of rating data. Motivated by our prior work on the investigation of user reviews that broadly appear in such sites, we hence propose a new recommender algorithm by fusing a self-supervised emoticon-integrated sentiment classification approach, by which the missing UserItem Rating Matrix can be substituted by the virtual ratings which are predicted by decomposing user reviews as given to the items. To test the algorithm's practical value, we have first identified the self-supervised sentiment classification's higher performance by comparing it with a supervised approach. Moreover, we conducted a statistic evaluation method to show the effectiveness of our recommender system on improving Chinese online video recommendations' accuracy.

Sentiment Analysis on Twitter about the Use of City Public Transportation Using Support Vector Machine Method

“…Figure 3 shows that the accuracy of the dataset C has the best accuracy values among the three existing datasets, amounting to 76.81%. In accordance with what is disclosed in the previous research (Zhang, Weishi, Ding, Chen, & Li, 2010) that the dataset C which has a balance composition of positive and negative data will result in a better classification since avoiding bias or deflection of weight word with a high number of occurrences in the negative data only or positive data only.…”

“…In research (Zhang, Weishi, Ding, Chen, & Li, 2010), it is said that the composition of the positive and negative of data need to be balanced to avoid any bias or weight deflection with a high number of occurrences in negative data only or positive data only. Therefore this test done to prove whether the situation also applies to the data of this research.…”

Improving Cold-Start Recommendations with Social-Media Trends and Reputations

“…Similarly, Jakob et al [9] identified and clustered movie topics using movie reviews and a ratings matrix to recommend movies based on the prediction of user-topic similarity. Zhang et al [10], obtained good results in online video recommendation with a framework that extracts a like/dislike rating from textual reviews and comments to minimize data sparsity.…”

Text-Based User-kNN: Measuring User Similarity Based on Text Reviews

“…Existing work [3,4] reports that measuring the similarity of users using the sentiment of their text reviews, instead of ratings, improves the accuracy of user-kNN. However, we argue that a sentiment-based approach does not fully address the similarity reflection problem since the reasons behind a sentiment of a review remain unexploited.…”