Mining key information of web pages: A method and its application

摘要

Web content mining aims to discover useful information and generate desired knowledge from a large amount of web pages. Key information, such as distinctive menu items, navigation indicators, which is embedded in web pages, can help classify the main contents of web pages and reflect certain taxonomy knowledge. Therefore, mining key information is significant in helping acquire domain knowledge and build catalogue classifiers. Current web content mining methods cannot mine such key information effectively. "Noise information" (such as advertisements) is a problem for the performance of web mining tasks. This paper proposes a method to extract key information out of web pages which contain noisy information. The method contains two steps: to extract a list of candidate key information, and then apply entropy measure to filter noisy information and discover key information. Experiment results show that this method is effective in discovering key information. With the discovered key information that reflects taxonomy knowledge, an application is developed to help ontology generation.