Recent Developments and Reviews in Sentimental Analysis for Big Data

Abstract- with the increasing need for understanding customer behavior and need for better buyer-seller relationships more than ever sentiment analysis has become one of the major tool in today’s time. The growing data and the need for faster computation efficient and more reliable processes of SA (sentiment analysis) are preferred and are in great demand.SA as a field of science has grown a lot from its earlier days. With the advent of big data practices, this paper focuses on processes followed in performing SA on big data and how big data tools and frameworks go along with sentiment analysis and it also highlights the gaps and suggests future works that should be explored.SA studies need to be expanded into providing better scalability and velocity along with reliability.

I. INTRODUCTION

The growing need for the powerful computational means and high speed analysis big data tools and frameworks gained recognition. Big data also enables companies to collect and analyze diverse and unstructured data which in past was mostly ignored. People realized the potential that different sources of data hold. Sentiment analysis (SA) which is basically the study of patterns and relationships that emerge from the data sources. This analysis helps businesses make informed predictions and access the people’s response, attitude or emotion towards a particular product, campaign, service, agenda and ideas over the internet in the form of text, audio or video. The resultant output can be divided into three categories positive, negative or neutral. These categories comprise of many names and slightly varied tasks, such as subjectivity analysis, opinion mining, opinion extraction, sentiment mining, affect analysis, customer complaint, emotion analysis, review analysis and review mining. Many techniques of sentiment analysis are developed which can be characterized into: Application based, which consists of stock market predictions, supply chain assistance ,business strategy analysis etc. ; fundamental approaches, including word-level sentiment disambiguation, sentence-level Sentiment Analysis, aspect-level Sentiment Analysis, bigram trigram and multigram SA, concept-level SA, multilingual SA and linguistic features analysis; and social aspect, which exploits the online content generation to analyze such inputs as epidemic spreading, emotion and responses towards events, campaigns and products.
Read More