Abstract
Abstract Several early studies had found that most users’ browsing behavior were concentrated in a minority of website(s). This phenomenon related with power-law and 80/20 rule, e.g, 80 percent of the users’ browsing behavior concentrated in 20 percent of websites, or 20 percent of the users account for 80 percent of the website’s flow. This paper proposes concentration index to measure the degree of users’ and websites’ concentration in the browsing behavior and discuss this issue further. Finally, I use the regression analysis to find out which variable(s) could influence the Gini index of page views, visits, duration. This research adopted one month (Oct. 2003) data to discuss. The findings are as follows: (4) The average visitor of each website was 36.2, the average visits of each website was 235.8, the average page views of each website was 1049.4, the average duration of each website was 6.08 min。 (5) The part of users：the average of Gini(Pageviews) is 0.7545, the average of Gini(Visits) is 0.561, the average of Gini(Duration) is 0.7776. The part of websites：the average of Gini(Pageviews) is 0.534, the average of Gini(Visits) is 0.4089, the average of Gini(Duration) is 0.6319。 (6) Regression analysis：the Gini’s indexes of page views, visits, duration are related with the website flow, e.g, the number of page vies, the frequency of visits, the time of duration and the number of user. Key word: Concentration, Concentration Index, Gini Index, Browsing Behavior