Study of Active-learning-based Theme Crawling System

This paper aims to design the active-learning-based theme crawling system and to realize the capacity of active learning of theme crawling by the use of semantic functions of ontology. The most conspicuous feature of this crawling system is the introduction of iterative process of ontology-incremental active learning. With the above circulating iterative process, the system can capture large quantities of theme-related web pages and the relatedness of the obtained web pages and the capturing rate are obviously superior to the crawling system without the ontology.