Predicting demographics features of the SNS by using Amalgamative Classification Algorithm (ACA)

  • Kotaiah Swamy Kakulla, Gouse Baig Mohammad, Ramu Kuchipudi


The internet has become unmanageably big and day by day it is increasing exponentially mainly through social media, blogs and reviews. Most of this information is written by various authors in different contexts. The availability of information put a challenge to researchers and information analysts to develop automated tools for analyzing such information. In this regard, Author Profiling is a popular technique attracted by several researchers to extract as much information as possible from the texts by analyzing author’s writing styles. Writing styles are many types such as messages in social networking sites such as facebook timeline messages, twitter posts etc. Based on the given social networking data of every user it is easy to predict the demographics features of the various users. In this paper, a new amalgamative classification algorithm (ACA) which classify the various types of users with different features and their ideas on latest trends. To improve the performance of the proposed system pruning is the technique adopted. The dataset used in this paper is facebook dataset for classification. Results show the performance of proposed system.