Automated detection and guessing without semantics of sender-receiver relations in computer-mediated discourses

Abstract

Automated tools for discourse analysis process tremendous amounts of computer-mediated discourses in a fast way. Assigning receivers to each message is an important step to answer the question "Who is communicating with whom" at any time. Direct addressing helps but is not used in every message. The focus of this paper is mainly split into two parts. First, automated detection and mapping of written receivers (or parts of them) to logged-in users. Second, if no receiver is written, automated receiver guessing without semantics. The architecture of the automated software is described in detail. As an applied example, the well-known and text-based chat system Internet Relay Chat (IRC) is used. An IRC discourse with 5605 messages is manually and automatically analyzed for comparing both approaches. Both - detection and guessing - are similarly well done as by hand.