New user transaction algorithm
一种新的用户事务算法

This study proposed a user business algorithm with double thresholds. This algorithm first acted according to the page number which the user visited to judge whether this user was the accidental user, and then the network topology and homepage lowest interest degree to judge whether the homepage appealed to the users. This method improved the data pretreatment process, and deleted the visit record which the accidental user caused, as well as the link pages and the pages that users were not interested in, produced one kind of effective visiting page sequence, namely double thresholds user business. This paper has proved the validity of the algorithm through an instance.