Ontology-Based Sentiment Analysis of Kazakh Sentences

摘要

Sentiment analysis one of the important and interesting task in natural languages. A number of resources and tools have been developed for sentiment analysis of English, Turkish, Russian and other languages. Unfortunately, there was no data and tools available for sentiment analysis in Kazakh. The Dictionary of Kazakh sentiment words has been created during the study. In this work we described the rule-based method using dictionary for sentiment analysis of texts in the Kazakh language, based on the morphological rules and ontological model. Ontological model for rule extraction that determines sentiment was built. Our rule based method achieves 83% accuracy for simple sentences.