Contrastive Opinion Summarization based on Blog Threads

Computer Engineering

Social media has become a vital medium for people to analyze others opinions while making decision. These opinions consist of both positive as well as negative opinions about that particular product/topic. There are large amount of opinions shared using blogs, forums or other opinion dedicated websites. Thus referring all those opinions is a challenging task for user. Existing opinion summarization work is limited to separating positive and negative opinions on basis of particular aspects or features. This paper gives a solution to this problem by generating a contrastive opinionated summary which highlight both positive as well negative opinion in form of pair so user can view both types of opinions at same time. For the computation of contrastive summary basic level similarity measures are used. Our goal would be to generate a contrasting opinion summary using advanced sentence similarity measures for better results. This facilitates direct comparison among the opinions in form of a contrastive summary which helps in quick decision making.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.