Personalized Marketing in Facebook using Hidden Markov Model

Abstract

In the current dynamic business world, each business is compelled to keep up with the state-of-the-art technologies to survive in the ever competitive market. With the dawn of the digital era, marketers were opened up to a brand new path of marketing strategies that allows them to not only manage their existing consumers but also to capture the world wide attention of potential consumers at a trivial cost. Social networks are now pronounced to be the next generation of consumer-centric interactions. This is proven further by the data released by popular social networks like Facebook regarding their consumer interactions. Big data analysis is another innovative technology concept used by businesses to recognize consumer buying patterns. Big data analysis has made path to web data mining. With the growth of social networks its profile holders have become the creators and distributors of business data. Due to this, social networks hold terabytes of raw data. With these massive advantages in line, web mining techniques and theory requires to be researched thoroughly to uncover the potential patterns and user access behaviors of these large amounts of online data. The major dilemma faced by researches in this field is the dynamic behavior of online data. To add to this dilemma the data is also quite unstructured. With these difficulties, the need to discover an efficient approach to mine unorganized and dynamic data in social networks is of utmost importance. This research aims to fill this research gap by introducing a framework that can be used to mine profile user data extracted from Facebook and use the knowledge in personalized marketing. This research paper explains framework build based on Hidden Markov Model to predict the next movement to handle the dynamic behavior of web data.