Age Group Prediction on Textual Data using Sentiment Analysis

Social media platforms provide a large amount of textual data covering various topics to explore opinions and emotions, hidden in the content using sentiment analysis. The consumer perspective on the quality and popularity of a product can be deduced from the product reviews, available at social media platforms by performing sentiment analysis. Sentiment analysis tells about the polarity of a sentence whether positive, negative or neutral. It can be used to predict personality, age and gender, based on writing style using feature extraction on the labeled training data sets. Understanding human emotions and opinions from text is a difficult task and to make it easier, sentiment analyzers are used. This paper proposes a method for prediction of age groups namely teenagers, adults and senior citizens from textual data collected from twitter and compares performance of different classifiers such as K-Nearest Neighbor (KNN), Multi-layer Perceptron (MLP), Decision tree, Random forest and Support Vector Machine (SVM), based on certain performance metrics like f-score, precision, recall and accuracy. One of the basic applications of this work can be for web readability analysis of resources, available on Internet.