Navigation in knowledge-based system for helpdesk based on FCA

Abstract

An important part of knowledge-based systems for helpdesk is a database of incidents that were solved in the past. A major part of the reported incidents is related to relatively small sets of repeated problems. It is important that these repeated problems are correctly recognized and their solutions are stored. In order to solve the new incident we can use the experience of solving a similar previously reported problem. Such a process is known as Case Based Reasoning (CBR). In practice however, it is not always immediately clear what the problem is about. It is due to the fact that different customer may describe the same problem in a different way. In this paper we describe an application of the helpdesk system which is based on Formal Concept Analysis (FCA), and enables to find analogous incidents by using navigation within the concept lattice. © Springer-Verlag Berlin Heidelberg 2007.