Applying Text Mining and Social Network Graphs to Analyze Online news of Shopping Websites

• Main Authors: Huang,Tzu-Yi, 黃姿倚
• Other Authors: Ma,Li-Ching
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/eerp2n

碩士 === 國立聯合大學 === 資訊管理學系碩士班 === 107 === With the development of the Internet era, e-commerce has grown rapidly. Many people choose to buy goods on the Internet. It is not only convenient but also fast. As long as customers can access the Internet, they can purchase goods through the online store no matter where they are. In recent years, the number of people using shopping websites to purchase goods has increased year by year, and online news and comments have greatly influenced consumers' decision to buy products. In such a highly competitive environment, a shopping website can use online news to analyze their strengths and weaknesses, and formulate relevant strategies to enhance competitive advantage. This study takes three famous shopping websites in Taiwan, including PChome, Yahoo Shopping, and Momo shopping, as examples to analyze related online news by text mining techniques and social network graphs. Frequent terms of online news for each shopping site and relationships among three shopping sites are analyzed and compared. Then, the Genetic algorithm is employed to calculate the coordinate of each frequent term, and the concept of social network graph is used to display those relationships graphically. Sentimental analysis is adopted to analyze the positive and negative tendencies of each term. In addition, this study uses machine learning methods, including decision tree, support vector machine and neural network, to analyze the impact of the content of on-line news on the stock price, which can be provided to investors for reference. The analysis results can help the managers of the shopping website understand the impressions and opinions of the website from customers’ viewpoints, and help them to adjust their business strategies and enhance their competitiveness.