Prediction of stock trend using transaction evaluations in Rakuten mobile phone market in Japan

• Main Authors: Pang-Yen Teng, 鄧邦彥
• Other Authors: Jun-Lin Lin
• Format: Others
• Language: zh-TW
• Online Access: http://ndltd.ncl.edu.tw/handle/6qkuz3

碩士 === 元智大學 === 資訊管理學系 === 103 === With the rapid expansion of the Internet, the volume of data on the Internet increases substantially. Meanwhile, mining the data on the Internet to support decision making has also become increasingly important. Due to its convenience, shopping online is now a major life style in our everyday life. The amount of transactional data from e-commerce websites grows quickly. In this study, how to utilize these transactional data from an auction website to predict stock trend is investigated. Traditionally, the stock trend was predicted using past stock-related data, such as stock price and stock transaction volume. In this study, we augment these data with transaction data from an online auction website. Our experimental results show that the prediction accuracy can be improved when the stock is in the up-trend category.