A Sentence Level Classification of Telugu News Document using Sentiment Analysis

Abstract

In recent years, sentiment analysis-based categorization in low-resource languages and regional languages has become a hot topic in natural language processing. Researchers are more interested in categorizing sentiment in Indian languages such as Hindi, Telugu, Tamil, Bengali, Malayalam, and others. To the best of our knowledge, no microscopic study on Indian languages has been published to yet due to a lack of annotated data. Using Telugu sentiment analysis, we presented a two-phase classification technique for Telugu news phrases in this work. It first recognizes subjectivity categorization, which categorizes statements as Positive, Negative, or Neutral. Sentiment Classification is the next step, which divides subjective statements into positive and negative categories. We get an accuracy of 81 percent for sentiment analysis categorization using this method.