论文部分内容阅读
用户行为分析是Web站点信息推荐中的重要方法,被广泛应用在该领域的诸多算法中.PageGather算法是其中有代表性的一种.旨在解决静态PageGather算法输入数据量过大、时间复杂度高的问题,使其更具实用性.通过引入渐进学习和分布的机制,给出了改进的算法PG+和PG++,并进行了实验分析.改进后,既保证了算法的等效性,又明显提高了效率.
User behavior analysis is an important method in Web site information recommendation and is widely used in many algorithms in this field.PageGather algorithm is one of the representative ones.It aims to solve the problem that the PageGather algorithm inputs too much data and time complexity High and make it more practical.Through the introduction of the mechanism of gradual learning and distribution, the improved algorithms PG + and PG ++ are given and analyzed experimentally.The improved algorithm not only guarantees the equivalence of the algorithm but also obviously Improve efficiency.