| Summary: | 碩士 === 中原大學 === 資訊管理研究所 === 98 === WOM (word-of-mouth) has a strong effect on consumer behaviors. Consumers often consider others' experiences through the Internet as suggestions before making a buying decision. However, too much information of WOM overwhelms consumers. Manually collecting WOM becomes a very time-consuming task. Therefore, this study proposes a restaurant WOM searching mechanism, which contains WOM collecting, document processing and adaptive suggesting functions. More specifically, the mechanism takes advantage of text-mining to analyze and extract articles in order to generate property’s weights for restaurants. This mechanism further recommends consumers according to their adaptive preferences. Experimental results show that our restaurant searching mechanism achieves a better performance than direct searching and blog searching based on the effective and efficient comparison criteria.
|
|---|
