A Comparison of Preference Elicitation Methods on the Internet Using Personalized Advertisement

• Main Authors: Yang, Chun-Cheng, 楊峻程
• Other Authors: Wen, Kuang-Wei
• Format: Others
• Language: zh-TW
• Published: 1998
• Online Access: http://ndltd.ncl.edu.tw/handle/51418846030959304313

碩士 === 國立中正大學 === 資訊管理研究所 === 86 === 　　The rapid growth of World Wide Wed and lnternet has brightened the prospective of lnternet commerce. Companies are increasingly emphasizing customer - directed production and marketing. Since the visitors of a web site are also potential custiomers of the products and services provide by the company, understanding the preferece of web surfers will provide useful inference to the propensity of potential customers. Today, with the availability of the lnternet technology, companies could seek to better understand their customers on the WWW. Currently, they could elicit preference via on - line questionnaire, or using "track analysis" －a method of observing viewer behavior on web site and analyzing the "traces" from logfiles recorded by web server to infer preference. The purpose of this study is to copare the effectiveness of these two methods using personalized advertisement. 　　In this study, we accomplish our objective using a laboratory experiment. Our results indicate the following. Fisrst, we revalidate that derived preferences using track analysis are close to true preference. Second, The traditional questionnaire method is better than the "track analysis" in overall advertisement effectiveness. However, for the significant product attributes, the two methods performed equally well.