The Designing and Prototyping of Dynamic Commodities Search Engine on the Internet - A Case Study on Electronic Bookstores in Taiwan

• Main Authors: Ya-Hong Hwang, 黃雅紅
• Other Authors: Prof . Chong-Yen Lee
• Format: Others
• Language: zh-TW
• Published: 2001
• Online Access: http://ndltd.ncl.edu.tw/handle/67842945505688277485

碩士 === 中國文化大學 === 資訊管理研究所 === 90 === This study concentrates on setting up the rules and analyzing the structure of shop-ping homepages from HTML documents to retrieve the text information of the specific commodities automatically; besides, it also provides the self-designing program to ex-plain the theory practically. There are no formal standard for those homepages on nowadays; in other words, it is hard for the customers to organize and gather informa-tion of commodities that were separated in different homepages, and customers have to repeat the same movement to get in and out of different homepages to collect the whole information. To simplify the searching procedure and to improve the efficiency of shop-ping on Internet, forming clear rules of retrieval and analyzing the structure of home-pages are very important. This study induces the analyzing in both the interrelated in-formation and also the distinction between the display of commodities from different shopping homepages; those information will be used in designing automatic analyze through the program. The most advanced integration technique still has long way to go. There are sev-eral search engines have been integrated between the shopping sites through the busi-ness cooperation. The customers can query the complete information of commodities only via one interface to make their decision; however, it can be worked only under the specific allied homepages that were under the same federation. But, there is still a prob-lem when it doesn’t work under the federation. Among the dozen most popular use of tables on shopping homepages, the charac-teristics of the display of tables that were already be analyzed and classified. There are four popular categories of the format to represent the method of information retrieval. First, there are multi-records in one table with only one title. Second, there are multi-records in one table, and each record has his title. Third, there is only one-record with the title in one table. Forth, there are multi tables for one record that with itself title but not all the information of commodities are in the same table. Depending on the original codes between these categories, the specific technique be developed to retrieve the shopping homepages dynamically. Finally, a experimental parsing system and a simple search engines that adapt the example of Taiwan electronic bookstores are being developed to test the proposed method.