Sentiment Analysis For Enhancing Business Process Decisions Using Machine Learning Techniques

Abstract

newline Human behavior is considerably prejudiced by their quirky sensitive state and opinions such newlineas attitude, feeling or sentiment. With the advancement and availability of internet, people newlinenow are able to share online their opinions, believes and feelings about services, products, newlinecurrent events, political issues or on any topic of their choice at any time. The huge amount newlineof data engendered by people on the web if anyhow can be exploited effectually may lead to newlineexpedient information and advantageous insights. The treasured perceptions gained can be newlinevital for business firms to understand requirements of improvement in their products or newlineservices, craft targeted marketing strategies and track economic shifts. newlineThe tricky question arises now that how to deal with this unstructured, unorganized and bulky newlinedata for gaining valuable insights. The studies in the field of machine learning and sentiment newlineanalysis are targeted towards delivering automated solutions for predicting and determining newlinethe opinions to get crisp information which can be exploited in real-world applications. This newlineresearch offers an intelligent decision system based on the framework of sentiment analysis newlinethat works based on the machine learning approach. It sustenance mixed-opinion text and newlinemultiword expression for sentiment analysis to determine the sentiment expressed. The newlineresearch is also focused towards estimation of aspects to which those sentiments are related newlineto providing precise information about liking and disliking of customers towards a product or newlinea service. It also helps the managers to take wise decision for a product launch, product design newlinechange, product feature change, and service quality related issues. newlineThe research utilizes publicly available datasets across two domains customer interaction newline(telemarketing data of bank customers) and customer review data (restaurants, electronic newlineproducts, and movies) to evaluate the intelligent decision system and sentiment analysis newlineframework for its accuracy and reliability. The sizable perf