Using Product Reviews from Twitter for Mining Insights to Predict Sales

Abstract: Tweeter, Facebook, Youtube are common terms now that everyone knows and most people are using them. People have social networking accounts. Social networking activities are a part of social life now days. Human beings interact with each other through social media. Most of the companies who provide services or product have their social accounts on social media to endorse their products or services. They often post online about new product or services. People write reviews about product experiences. These reviews are large in number. We can utilize this data to help a vendor to make intelligent business decisions. This paper presents a system used to mine the predictions of a product. Tweeter data was made use of. A tweeter user posts tweets about a product or service. These tweets were used as database and processed using DynamicLMClassifier algorithm. Based upon the classification, tweets are stored into categories. Categories were negative, positive and neutral. This classification was very helpful to take business decisions.

Keywords: Data mining, Business Intelligence, DynamicLMClassifier, Review Mining.