Making Sense of CRM Messages: an Interactive Toolset

CRM managers are frequently overloaded with large number of text messages from their customers, and making sense of those messages is a difficult task. We introduce an interactive toolset that can facilitate the exploration of CRM data semi-automatically. This toolset is based on the state of the art text processing technologies and allows CRM managers to discover interactively the re-occurring issues and trends in vast amounts of customer messages. Our design of the toolset is based on an iterative sequence of steps: 1) identifying descriptive terms, 2) identifying semantic relationships between the terms, and 3) grouping CRM messages into clusters of related issues. This paper presents a prototype implementation of the toolset, the justification for the specific toolset design choices, and an ongoing field study with CRM managers at a computer customer support center in a large university.