Improving Search Results with Lightweight Semantic Search

The goal of each search service is to yield the most relevant results on a given query. Traditional full-text search is not enough and many approaches to improve search rankings are adopted. In this paper we propose a method of combined search query scoring computation leveraging lightweight semantics represented by metadata related to searchable content. It extends state-of-the-art approaches at both indexing and searching stage. We discuss two approaches of so-called concept scoring computation in order to capture different properties of available metadata.