Twitter Sentimental Analysis using Natural Language Processing and Dynamic Architecture of Artificial Neural Networks

Abstract

In this day and age our society is bending towards the use of various online platforms for gathering feedbacks, reviews, opinions. People are being more active on various networking sites and provide their comments on various products. Furthermore, there are many irrelevant feedbacks and comments that don’t provide any benefit in terms of business or social upbringing of the society. It would engage a lot of time of the business associates and other team to come up with the useful feedback. Hence in this state of affairs we introduce sentimental analysis for solving this issue. To process all the data, we need different algorithms [12] and we here are using natural processing language. In initial stage we have so many feedbacks and comments that need to be structured and hence we propose this algorithm that is efficient and it would save a lot time and effort. We would be dissecting the tweets in three degrees: positive, negative and neutral. By doing so we would be able to make out the positive and negative comments. We are focusing on various deep learning techniques which we have used in our paper to provide an effective analysis. We obtain an accuracy of 86.25% using natural processing language [7]. We discovered that sentimental analysis is considerate in analyzing the tweets as well as the seriousness of each tweet. Sentimental Analysis is exceptional way of analyzing the vast unstructured data and helps to perform compressed study on the feedbacks and review. This paper calls attention to the purpose of extracting the useful and irrelevant comments and natural processing language assist us for gaining optimum solution.