Industrial symbiosis (IS) is a subdiscipline of Industrial Ecology that aims to bring together companies from different sectors to share resources, namely energy, materials, and water. The main goal of IS is to improve resource (materials, waste, energy) efficiency and lead to mutual environmental, financial and social benefits to participants. In this paper we present a semantic approach for IS input/output matching. This approach is based on knowledge modelling and ontologies. Ontologies are used to model all resources - waste, water, energy - along with details about their composition, characteristics (chemical and physical) and tacit knowledge about their flow. The input/output matching algorithm presented enables the valorisation of resources through industrial symbiosis networks. (c) 2014 Elsevier Ltd. All rights reserved.