Semantic Enrichment in Knowledge Repositories

Abstract

Among various knowledge management initiatives, the creation of knowledge repositories has emerged as a prevalent approach in current knowledge management practices. In order to transfer tacit knowledge from individuals to a knowledge repository, organizations usually deploy community-based discussion forums. Typically, discussions among participants are organized into reply-replied structures, and reply semantic relationships among these discussion documents exist either explicitly or implicitly. Such relationships, once discovered or annotated in the knowledge repository, can facilitate subsequent knowledge navigation by providing a novel and more semantic mechanism and can support other organizational knowledge management activities (e.g., construction of expert networks). In this study, we propose a preliminary taxonomy of reply semantic relationships for discussion documents organized in reply-replied structures and develop a Semantic Enrichment between Knowledge-sharing documents (SEEK) technique that automatically annotates semantic relationships between reply pairs of documents. Specifically, we propose and evaluate six different feature models that combine keyword features, part-of-speech statistic features, and/or text statistic features.