SEMO: Searching Majority Opinions on Movies using SNS QA Threads

Many people seek majority opinions by searching for question-answers that are uploaded by others or uploading their own questions on social media sites. However, people have to read through a large number of documents returned by search services to find the majority opinions. Moreover, even when users upload questions on social media sites, they cannot immediately obtain answers. To address these problems, we present Searching Majority Opinions System (SEMO), a novel majority opinion-based search system that uses QA threads uploaded on SNS and cQA websites. SEMO returns entities based on majority opinions for opinion-finding queries in real time. We also tackled a data sparsity problem using a novel query component expansion approach. To prove SEMO's usefulness in finding majority opinions, we implemented a prototype of SEMO for the movie domain. We believe that our method can cause a paradigm shift in opinion-finding query search and help people make decisions. SEMO is available at http://semo.korea.ac.kr/