Sentiment Mining of Community Development Program Evaluation Based on Social Media

Abstract

It is crucial to support community-oriented services for youth awareness in the social media with knowledge extraction, which would be useful for both government agencies and community group of interest for program evaluation. This work provided to formulate effective evaluation on community development program and addressing them to a correct action. By using classification based SVM, evaluation of the achievement level conducted in both quantitative and qualitative analysis, particularly to conclude which activities has high success rate. By using social media based activities, this study searched the sentiment analysis from every activities comments based on their tweet. First, we kicked off preprocessing stage, reducing feature space by using principle of component analysis and estimate parameters for classification purposes. Second, we modeled activity classification by using support vector machine. At last, set term score by calculating term frequency, which combined with term sentiment scores based on lexicon.The result shows that models provided sentiment summarization that point out the success level of positive sentiment.