Mining electronic negotiation messages – an exploratory evaluation and first steps

Analysing negotiation communication in an electronic setting is a challenging task. During recent years, many researchers have tried to explain the differences between successful and failed negotiation applying a great variety of approaches, ranging from linguistic analyses over phase identification to psychological analyses of the discourse during a negotiation. The aim of this paper is to provide the foundation to analyse negotiation communication with a machine learning approach. To this end, a review of existing and related research is given, the applicability of machine learning techniques on electronic negotiations is discussed and, in a first step of the process, different methods of feature selection are reviewed to find the most applicable technique in the setting given.