Sentimental Space Based Analysis of User Personalized Sentiments

Abstract

With the development of social media, online documents such as the comments of news articles, blogs and microblogs have received great attention, and the sentiment analysis via online documents has become one popular research area. This paper focuses on establishing user sentimental space obtained from online documents to analyze user's personalized sentiments, which aims to identify user's sentimental feature. Affection, sentiment and attributes of user are firstly employed to build user's personalized sentimental space. Then, the general constrains of user sentiments space are proposed to calculate user's personality. And finally we seek out sentimental leaders who paly pivotal role in the leading public opinions. Our works can give some suggestions for decision makers when urgent event happen.