Application of Web Mining to Network User’s Articles - an Evidence form mobile 01

碩士 === 國立臺北大學 === 企業管理學系 === 99 === This research is related to mining the characteristic value of the article, analyzing the relatedness of characteristic values, and comparing article trend with the movie box office sales in Taiwan. The study period is during December 24, 2009 to April 20, 2010, a...

Full description

Bibliographic Details
Main Authors: Liu, Chienlee, 劉千里
Other Authors: Goo, Yeongjia
Format: Others
Language:zh-TW
Published: 2011
Online Access:http://ndltd.ncl.edu.tw/handle/32735838774103664442
Description
Summary:碩士 === 國立臺北大學 === 企業管理學系 === 99 === This research is related to mining the characteristic value of the article, analyzing the relatedness of characteristic values, and comparing article trend with the movie box office sales in Taiwan. The study period is during December 24, 2009 to April 20, 2010, and a total of 814 discussions about Bodyguards and Assassins, MONGA, and Shutter Island on the forum of mobile 01. Mining methods used frequency analysis, coefficient correlation and weighting method of characteristic value in sequence. The result of study shows that we can discriminate the article trend rapidly via observing the characteristic value and related intensity of characteristic value. Finally, through the evidence of weighting method of characteristic value we can find that article trend of the forum and the actual amount of Taiwan's total sales have a high degree of correlation.