Improve the Sentiment of Bengali Language Texts with Stopword Removal

Many sentiment analysis research studies have been conducted, mainly based on the English language. To clean up the English language dataset, several applications have been developed with built-in functions to facilitate the research work. One such function is the Stopword function. It is frequently used to eliminate unnecessary words from sentences to preprocess datasets in various languages. Similarly, research is being carried out to analyze sentiments using the Bengali language. Bengali language is also filled with unnecessary and noisy words that do not express feelings. The objective of this research is to demonstrate that the removal of stopwords can significantly enhance the accuracy of sentiment analysis in Bengali language texts. In this paper, a Stopword function was implemented to remove unwanted words from a dataset of Bengali sentences, and different machine learning algorithms were used for classification. The outputs were compared before and after the implementation of the Stopword function, resulting break a hand accuracy.