Developing a Sentiment Analysis Model for Online Brand Management and Optimization

Online brand management is very crucial in today's world with prominent relevance if customer reviews. One of the prime brand optimization techniques include the major fundamental application of Natural Language Processing (NLP), i.e. sentiment analysis. This is prominent when there is abundance of feedback and reviews of customer utilizing the brand services. This paper focuses on the approach of studying and analysing the brand reputation in different business verticals and applying the NLP techniques and classifier models to attain the efficiency. The classifier models that have been trained in this research includes the linear regression, Random Forest Classifier and Gradient Boost Classifier. The two major verticals that have been considered are, namely Transport & Logistics and Finance & Insurance, with reference to diverse datasets being trained on the models. The maximum f1 score of 97.4 has been achieved on the Gradient Boost Classifier making it the most effective model in this research.