Social Media Data Aggregation and Mining for Internet-Scale Customer Relationship Management

Abstract

Industry and government organisations across the world increasingly use social media for Customer Relationship Management (CRM) activities that include communicating with customers about their services/products and distilling customer opinions for assessing their effectiveness. In this paper, we explore the exploitation of social media for CRM, and describe the development and functionality of VIZIE - a social media monitoring system designed to help media monitors to distill the perception of potential and actual customers of the value of existing products and services, identify and correct customer misinformation, discover how to better meeting customer expectations, and (if needed) communicate directly with the customers by using the customers' preferred social media forum. VIZIE monitors the most popular social media networks and mines their content in real time independently of topical domain and social media source/type. VIZIE achieves these by integrating information from multiple social media sources to create a broader, more complete context compared individual social media posting about specific topics of interest to a VIZIE user. VIZIE also provides the ability to automatically create overviews and summaries of any such topic, permits user queries, and supports drill down to individual social media posting. VIZIE accomplishes these by using a variety of natural language processing and information retrieval techniques that are specifically selected for (or tailored to deliver) real-time responsiveness. We provide an overview of the real-world requirements that informed the design of this system that has been used by more than twenty organisations and describe an overview of its architecture, the data analysis methods employed in its implementation, and their orchestration for achieving real-time response at an internet-wide big social data analysis scope.