Negotiation Closer Bot: Negotiation Outcome Predictor based on Sentiment Analysis

On e-commerce platforms like OLX, Craigslist, etc. users can sell and buy a variety of products and also chat with the buyer for additional information such as pictures, description and more importantly for negotiating the prices. This process of negotiation is tedious for all users and some simply do not know how to negotiate. To tackle this, a system that assists negotiations for the user would solve the problem. Therefore, we have created a web application platform that allows users to list products to buy and sell products with an added feature of a negotiation support system that assists users in the negotiation process by predicting whether the negotiation is going to end in success or failure. Users can trigger it whenever they deem it necessary by clicking on the ‘Get Recommendation’ button. This system will read the chat using functions of Natural Language Processing Toolkit (NLTK) library followed by classification using classifier. After testing four classifiers namely, Naïve Bayes, Descisioon Tree, Support Vector Machine (SVM) and Random Forest, we selected SVM as it gave the highest accuracy.