Identify interest and attitude of academy students based on confession posts analysis

In our research, we propose an text analysis methods to mining AOF confession post in a year to find out the issues that students is interesting and assessing the feelings. As a result, the model automatically extract students’ interest according to 3 interest domain and built an automatic sentiments recognition model with 52% F-score accuracy. This research is the potential step to build an automated system of assessing young people’s perspectives during the social networking boom.