Semantic distance acquisition in SemaCS

Abstract

Search functionality and technology is a growing area of research. However, simple search approaches are still frequently used. A simple keyword or thesauri-based search is efficient and can be easily scaled. However, keyword-based search cannot be used to infer what may or may not be relevant to the user and thesauri, or any other expert generated model, is expensive to produce and tends to be of limited applicability. Semantic Component Selection (SemaCS) approach is not tied to any specific domain and does not rely on expert input. SemaCS is based on actual data and statistical semantic distances between words. Information on semantic distances is used for searching and for automated generation of domain model taxonomy. This paper presents SemaCS's means of acquiring these semantic distances - mNGD (2) - and its initial evaluation.