The role of corpus linguistics in sentiment analysis of Persian texts, case study: a Farsi news agency website

Abstract

The current article aims to improve a linguistic model in the sentiment analysis of a Persian news agency website. After investigating many computational problems and shortages in the field of computational linguistics, we could see that the main problem of computational linguists could be found in linguistics, not in computational sciences. Presenting a model can lead to the management of uncertainty of semantic and sentiment analysis of Persian documents. The integration of systems that operate in the field can result in considerable developmental growth in smart systems of sentiment analysis of the Persian language in a way that could reduce complexities in the Persian language. Moreover, the presence of a comprehensive model can facilitate the generation of smart systems of sentiment analysis in text mining. First, we collected existing models for text mining of sentiment analysis and tried to suggest a model as a general principle. The obtained model will help for information management and planning of text mining systems in computational linguistics and shows the shortages of Persian language natural processing in line with the automation of sentiment analysis.