This paper proposed a Web log mining algorithm named PHIMA(public hotspot information mining algorithm), which based on area-channel to mine public hotspot information accurately and efficiently. On the analysis of the applicability of present Web log mining algorithms on e-government Web log mining, the algorithm used area-channel for mining public hotspot information. According to the conception, this paper proposed a user access matrix based preferred public hotspot information mining algorithm. PHIMA was proved accurate and scalable in the experiments. It's suitable for application in e-government, such as to optimize Web site or to provide decision support for policy-makers.