Research of searching strategy in topic crawler using dynamical particle swarm optimization
基于动力粒子群算法的网络蜘蛛搜索策略研究*

Traditional topic crawler,which uses monistic searching strategy,may cause the problems of topic drift,not utilizing structural information and being easy to miss searching direction.This paper proposed a new heuristic searching algorithm based on dynamical PSO,which considered the characteristics of Web structure and synthesized the advantage of linkage's immediate value and future value,especially could dynamically adjust the weight between the two rewards online.The kind of crawler was adaptive.The experiments show that this algorithm had better performance in recall rate and precision compared with some traditional algorithms.