K-means and Support Vector Machines for Thai Language IT News Retrieval System

The proliferation of information technology news websites on the internet has resulted in an abundance of content that can be quickly and easily published. There are numerous websites in Thailand that offer information technology news, which provides great benefits to both students and IT employees in the country. These websites are valuable resources for keeping up with the latest technological advancements and for sharing knowledge among IT professionals. However, searching through this massive amount of material to find relevant news in the Thai language can be a challenging task. Traditional search engine approaches rely on general keyword searches to locate pertinent information, but this method can prove problematic when dealing with extensive content in the Thai language. In response, this paper proposes an information retrieval system designed and developed specifically for Thai-language information technology news. The system employs the K-means algorithm and support vector machine for classification, serving as a pre-processing step for text classification used in the indexing process of the information technology news retrieval system. Experimental results indicate a recall value of 68% and precision of 95%. The F-measure performance was found to be 79%.