Real World Applications of Machine Learning Techniques over Large Mobile Subscriber Datasets

Abstract Communication Service Providers (CSPs) are in a unique position to utilize theirvast transactional data assets generated from interactions of subscribers with net-work elements as well as with other subscribers. CSPs could leverage its dataassets for a gamut of applications such as service personalization, predictive offermanagement, loyalty management, revenue forecasting, network capacity plan-ning, product bundle optimization and churn management to gain significant com-petitive advantage. However, due to the sheer data volume, variety, velocity andveracity of mobile subscriber datasets, sophisticated data analytics techniques andframeworks are necessary to derive actionable insights in a useable timeframe. Inthis paper, we describe our journey from a relational database management system(RDBMS) based campaign management solution which allowed data scientistsand marketers to use hand-written rules for service personalization and targetedpromotions to a distributed Big Data Analytics platform, capable of performinglarge scale machine learning and data mining to deliver real time service person-alization, predictive modelling and product optimization. Our work involves acareful blend of technology, processes and best practices, which facilitate man-machine collaboration and continuous experimentation to derive measurable eco-nomic value from data. Our platform has a reach of more than 500 million mobilesubscribers worldwide, delivering over 1 billion personalized recommendationsannually, processing a total data volume of 64 Petabytes, corresponding to 8.5trillion events.