The organization of an effective aspect-based oppinion mining system

Abstract

Social platforms are currently used as independent tools where people exchange information about their personal life and share common interests. On the other hand, these people use the Web to browse and find other information, e.g. an interesting destination for holidays, a doctor, a movie, a mobile phone, etc. Even with the emergence of the Web 2.0 - where a user can benefit from other users’ experience on the Web, users are still constrained to combine and synthesize themselves information issued from social platforms: read through all of the reviews, check if a user can be trusted or not etc. In this work, we investigate how to build a system that can help reducing the effort needed to synthetize opinions about an item and present it in a user friendly way with new visual metaphors. We focus on the general approach and describe the main steps for building such a framework.

Keywords

• opinion mining
• sentiment analysis
• aspect-based opinion mining
• aspect extraction
• summarization