Sentiment analysis to predict Bombay stock exchange using artificial neural network

Penetration of media in financial market has changed the way of doing business, media, one of the most growing medium to exchange ideas & thoughts, which plays an important role in influencing thoughts of the investors which in turn affects market transactions figures and index values. Thus machine learning and sentiment analysis can be of great help in deducing the mood and psychology of people which affects the market and thus can help us to predict the actual statistics. In this research work, sentiment analysis was formulated on data from social media which is classified using classification algorithm of machine learning. The classified data is analysed to calculate the net mood of the comments. These comments are classified into four classes´ namely happy, hope, sad, disappointing. The net relative mood of all the classes per day is used as input for artificial neural network (ANN) to be trained for data of n days and their respective change in index value on each day. This trained network is finally used to predict the vector of Bombay Stock Exchange index value for (n+1)^th days.