Poisson distribution-based page updating prediction strategy

A kind of method predicting the updating period of page is proposed, using the Poisson distribution-based model suggested by Brewington in studying page updating of search engine. This method calculates the expectation of page updating period with Newton iteration method according to historical statistical data then applies it in the subsequent updating strategy, also revises it in the practice in order to approach the real value; its goal is to get a good balance between freshness and resource utilization. Compared to traditional grouping-by-frequency method, algorithm complexity reduces significantly.