Survey on Text-Based Sentiment Analysis of Bengali Language

Digital contents in Bengali text are increasing with the advancements of World Wide Web (WWW) to express the sentiment of content originators from mixed perspectives. These high voluminous unstructured web contents can be utilized to build smarter tools to assist people through Natural Language Processing (NLP). Though Bengali NLP tools are still inadequate due to its inherent complexities, research on Sentiment Analysis in Bengali is flourishing as a challenging area and is getting researcher’s attention at a rapid pace. To do research in this area, researchers spend a lot of their time to investigate previously published works to understand the advancements for further enhancement; one of the most stressful and challenging parts of research. In this regard, we are going to demonstrate domain specific available pre-processing steps along with existing research methodologies, respective datasets and evaluation metrics of notable works related to sentiment analysis of Bengali language as our research finding. Moreover, we also incorporate a series of directions for future researchers to augment their research in this exciting domain.