Analyzing the Online Regularities of Users in Social Network via ”PTT” Discussion Group

[[abstract]]In this paper, the periodic regularities of user online in social network are evaluated with the number of user postings according to different periodic intervals. First of all, users' postings in social network are downloaded by web agent automatically. Secondly, the details of those postings, such as poster's name, post-time, responser's name, and its response-time, are extracted and collected via text mining techniques. Thirdly, the number of postings are accumulated according to three different periodic intervals, 24 hours/per day, 7 days/per week, and 12 month/per year. Experimental resources were downloaded from one social network, named "PTT", including four discussion groups as "anime", "stocks", "NBA" and "expectant-mother". Experimental results revealed that these users had special surfing habits according to different periodic intervals. These analyses of user online in social network provides important information for researching on internet addiction and seeking appropriate time-intervals for advertisement