| Summary: | 碩士 === 逢甲大學 === 應用數學系 === 107 === The community and self-media tendency grows up in recent years. People discuss social issues in an anonymous way through a social media community. Anonymity makes people willing to express their views on different issues and provide a different angle of thought. After a while, this type of text data has accumulated significantly on any issue. However, will the government policy promotion will be impacted by these public opinions ? That’s the main purpose of this study. We chose “Taiwan Plastic Restriction” as our topic, and the sample is from the Gossiping Board of PTT. PTT is the biggest Bulletin Board System (BBS) in Taiwan. We used the Sentiment Word Analysis to de-scribe people’s attitude to the policy and adopted the Word Frequency Method to find out the hottest vocabularies during the discussion. Then, we utilized the Word Association Method to find the words that highly associated with the policy. Finally, we used topic modeling to explore the 16 latent topics to build Thematic Networks by the Latent Dirichlet Allocation (LDA). We found the positive response rate to the policy is high. Both of the authors’ and responders’ emotions are related to mutually. Based on this research, we expect to provide some useful suggestions to enhance the community of government policy promotion.
|
|---|
