Customer Feedback Analysis using Collocations

Today's ERP and CRM systems provide companies with nearly unlimited possibilities for collecting data concerning their customers. More and more of these data are more or less unstructured textual data. A good example of this type of data is customer feedback, which can potentially be used to improve customer satisfaction. However, merely getting an overview of what lies in an unstructured mass of text is an extremely challenging task. This is the topic of the field of computational linguistics. Collocation analysis, one of the tools emerging from this field, is a tool developed for this task in particular. In this paper, we use the collocation analysis to study a text corpora consisting of 64,806 pieces of customer feedback collected through a case company's online customer portal. Collocation analysis is shown to be a very useful tool for exploratory analysis of highly unstructured customer feedback.