Abstract:

Short documents containing minimal text including business-related terms and figures are prevalent in business contexts but challenging to retrieve by end users in information retrieval systems. These documents often lack contextual information crucial for effective query matching. This paper aims to address this gap between users' needs and short documents by introducing a taxonomy model and query expansion approach. The proposed taxonomy is tailored to capture business information needs, leveraging semantic metadata features. Initial experiments conducted with an open document collection have shown promising results.