| Summary: | 碩士 === 元智大學 === 資訊管理學系 === 104 === Under the impact of the declining birthrate, the high education in Taiwan is faced the problems of recruit students. Now a day, because the prevalence of Internet enables people to more easily get messages and share with each other on the Internet. Therefore, application and master online reviews is good reference value for business a school. Especially, extract topic and sentiment of reviews, positive and negative semantic orientation, the term strength level, all of these are the focus of the text mining analysis. In this study, we use the web crawler to automatically retrieve online reviews of high education. Second, we separate the data into based on sentence level and based on document level. Third, we respectively for the sentences and document use Heuristic N-Phrase Rules to extraction topic and sentiment to effective analysis of university evaluation. This study would focus on seven topic of university, including school resources, graduation development, life function, tuition, appreciation, overall performance, select school situation, then give the positive and negative opinion. Through data compilation and correspondence analysis propose each school advantage, disadvantage and characteristics. This study give prospective students as a reference and let school operator as reference for decision-making.
|
|---|
