Sentiment Analysis using Ensemble Classifier for Entrepreneurs based on Twitter Analytics

When committing the time and resources to a startup idea, entrepreneurs must consider several aspects. Analyzing the thoughts of people about emergent businesses from popular social networking platforms can contribute positively in the decision making process of investors. Statistical methods are widely experienced in generation of business insights and estimation of future growth by determining a general trend in opinion of users. However, intelligent computing for efficient and affective sentiment analysis is still challenging and capture the interest of researchers. In this paper, we have aimed at using natural language processing to understand the attitude of twitter users about innovative business ideas and latest technologies in their surroundings. We have proposed an ensemble classifier to allow users to investigate the sentiment associated with a topic of interest in real time using data visualization. The objective of this research is to explore the abilities of machine learning algorithms in producing maximum accuracy on translating raw tweets into useful judgements through sentiment analysis. Our proposed application classifies positive and negative emotions taken from sentiment140 dataset. To assess the output of ensemble classifier, we have trained Support Vector Machine (SVM), Naive Bayes, and Long Short-Term Memory (LSTM) neural networks on a defined set of features. While, experimental results show that our methodology is more accurate than other known methods and may be used by entrepreneurs for improving business analytics. This paper also proposes a comparison among the current methodologies of sentiment analysis using artificial intelligence.