An intelligent agent-based system for multilingual financial news digest

Nowadays, online financial news from different sources is widely available on the Internet and there exists systems to help investors extracting and analyzing the financial news. However, many of these systems present news articles in a way without categorization and do not provide enough query options to search or get the specific aspect of news that they want. In this paper, we extend our previous work to develop an intelligent agent-based system for multi-lingual news extraction. We adopt a document categorization approach based on fuzzy keyword classification. The system applies fuzzy clustering to obtain a classification of keywords by concepts of the category. A category profile is developed and worked as a searching interface for document browsing. Experimental results show that the proposed Categorize News Agent is capable of categorizing news documents with a reasonable rate of accuracy.