DOTNET PROJECT
Mining Users Trust
From E-Commerce Reviews Based on Sentiment Similarity Analysis
Abstract:
Consumers' reviews in
E-commerce systems are usually treated as the important resources that reflect
user's experience, feelings, and willingness to purchase items. All this
information may involve consumers' views on things that can express interest,
sentiments, and opinions. Many kinds of research have shown that people are
more likely to trust each other with the same attitude toward similar things.
In this paper, we consider seeking and accepting sentiments and suggestions in
E-commerce systems somewhat implies a form of trust between consumers during
shopping. Following this view of point, an E-commerce system reviews mining
oriented sentiment similarity analysis approach is put forward to exploring
users' similarity and their trust. We divide the trust into two categories,
namely direct trust, and propagation of trust, which represents a trust
relationship between two individuals. The direct trust degree is obtained from
sentiment similarity, and we present an entity-sentiment word pair mining
method for similarity feature extraction. The propagation of trust is
calculated according to the transitivity feature. Using the proposed trust
representation model, we use the shortest path to describe the tightness of
trust and put forward an improved shortest path algorithm to figure out the
propagation trust relationship between users. A large-scale E-commerce website
reviews dataset is collected to examine the accuracy of the algorithms and
feasibility of the models. The experimental results indicate that the sentiment
similarity analysis can be an efficient method to find trust between users in
E-commerce systems.
No comments:
Post a Comment