Intelligence Gathering from Online NEWS Documents

Web documents are increasingly expanding everyday, which inhibits the practice of surfing the contents, especially news articles of one´s own interest. An automatic information retrieval (IR) system specially designed for retrieving Web-based NEWS articles of one´s personal interest would be of much use. This paper proposes a web-based focused IR system for collecting news articles about any chosen celebrity thereby reducing their precious time over surfing the net. A static domain-specific ontological framework about the focus of retrieval is fed to the system. The implemented system retrieves the focused information from the tagged news corpus by automatically classifying into various categories based on certain heuristics. The classified corpus is then subjected to information filtration by which the relevant semantic information present in the article is organized and projected as output. Precision and recall statistics are tabulated for politics and sports domain. By feeding alternate domain ontology, domain independent news information retrieval shall be assured.