摘要
为了有利于对网络用户实行个性化服务,采用先对服务器记录用户采用赋权值距离算法进行聚类,然后对各类缩小的用户群体采用BQ-tree树算法进行用户频繁浏览模式挖掘.仿真结果表明,整个算法在保证挖掘效果的同时,比以往的Apriori算法、FP-growth算法更节省时间,且挖掘结果能有效地对用户提供个性化服务.
For the convenience of individualization service for net users, the registered users of the server were first clustered with weighted distance assignment algorithm, then BQ-tree algorithm was used formining the user＇s frequent browsing patterns for various reduced users set. Simulation results showed that the whole algorithm was more effective and quicker than Apriori and FP-growth algorithm while it guaranteed the effect of mining. Further, the mining results could more effectively provide individualization service for the users.
出处
《兰州理工大学学报》
CAS
北大核心
2007年第6期72-76,共5页
Journal of Lanzhou University of Technology
基金
甘肃省科技攻关项目(2GS052-A52-003-11)
