Semantic search for matching user requests with profiled enterprises

Semantic search is an important approach that promises significant improvements for customers to identify products of their interest. To perform semantic search, enterprises need to publish semantically enriched descriptions of their offered goods and services; then a customer expresses his/her request, in an easy Google like fashion, by providing a list of desired features. If enterprise offerings and customer requests are based on the same vocabulary (i.e., ontology), they can be semantically matched by using advanced semantic methods. In this paper, we propose an ontology-based method aimed at finding the best matches between a user request and the services offered by different enterprises. We assume that in a given business ecosystem (in the paper, as an example, the tourism sector) a group of SMEs agree on the adoption of a reference ontology, used to build the company profiles on the basis of the offered services. Accordingly, a user request, represented by a set of desired features, is expressed in terms of the reference ontology terminology (i.e., concepts). In this paper, we illustrate SemSim, a method used to collectively search the SME profiles to identify the services that match at best the user request. SemSim is based on the well-known information content approach used to evaluate the semantic similarity between concepts. The experimental results show that our proposal performs better than some of the most representative similarity search methods proposed in the literature.

► We propose SemSim, an ontology-based similarity method for matching user requests against available enterprise services.
► SemSim is based on the well-known information content approach used to evaluate the semantic similarity between concepts.
► SemSim computes the similarity between ontology-based feature vectors used to represent requests and services description.