Combining Web usage and content mining to improve transaction identification
结合内容挖掘的事务识别改进

For the traditional methods of transaction identification couldn't properly meet the need of pattern identification,a novel improved one aiming at overcoming these drawbacks was proposed.The new algorithm which combines content data of web pages,and applies document clustering algorithm in this process,could not only take the user browsing manners and tendency of information needs into account,but also avoid the influence of outside factor such as buffer and network latency.As a result,the algorithm after improvement could acquire user's browsing pattern more exactly.