Mobile Portal Personalization using Real-Time Prediction Analytics in order to help Communication Service Providers (CSPs) in customized offerings to users
Submitted by Gitali Halder on Jun 30, 2012
With the phenomenal growth of mobile multimedia applications across various device vendor-controlled App Stores, Communication Service Providers (CSP) are immensely challenged to identify new revenue streams due to the gradual cannibalization of its voice and messaging revenues by next generation social media players from the Internet community. Our aim is to create a real-time solution to enable CSPs to offer personalized service to its customers based on subscriber’s interests, behavior, location, demographics and service consumption in order to reduce customer churn and increase customer loyalty by strategically positioning themselves in this Web-connected social media world.
The Mobile Portal Personalization is a recommendation framework based on subscriber’s usage information covering key use cases including surveys, deals, alerts, content-push etc, subscription details, location and demographic information though leveraging Vertica and R, thus enabling CSP to provide personalized service-offerings to its customers. The prediction framework constitutes many filters on variable selection and makes use of LM/GLM models for different types of user activities.
Speaker1: Jyotirmay Nag is currently an Analytics Consultant at Hewlett-Packard. He has 2 years of experience in Data Mining and Predictive Modeling in the areas of HR, Marketing, Sales, Finance etc. Prior to Hewlett-Packard he finished his Master of Statistics from Indian Institute of Technology, Kanpur.
Speaker2: Gaurav Sanghvi currently holds the position of Project Manager with the Strategy team of HP Global Analytics. He has over 7 years of experience in the areas of Analytics, Strategy Planning & Consulting, Market Intelligence, Competitive Intelligence and related areas across verticals like Technology, Banking and Financial Services. Gaurav received his Masters in Finance from ICFAI university.
Speaker3: Abhisek Saha is currently a Statistical and Data mining Consultant at Hewlett Packard Analytics. He has 5 years of experience in HP. Over these years he’s worked on many research-oriented projects in the areas of Predictive modeling, conjoint analysis, affinity analysis, text-mining, Bayesian modeling etc. He holds a Master in Statistics and a Bachelor of Statistic degree from Indian Statistical Institute. His works were presented before in ICADABAI 2011, ASONAM 2010 etc.