Abstract
Opinion mining and sentiment analysis have become popular in linguistic resource rich languages. Opinions for such analysis are drawn from many forms of freely available online/electronic sources, such as websites, blogs, news re-ports and product reviews. But attention received by less resourced languages is significantly less. This is because the success of any opinion mining algorithm depends on the availability of resources, such as special lexicon and WordNet type tools. In this research, we implemented a less complicated but an effective approach that could be used to classify comments in less resourced languages. We experimented the approach for use with Sinhala Language where no such opinion mining or sentiment analysis has been carried out until this day. Our algorithm gives significantly promising results for analyzing sentiments in Sinhala for the first time.
