This paper introduces HORUS (Human-readable Ontology Reasoner Unit System), a configurable reasoner which provides the user the motivations for every inferred knowledge in the context of a reasoning process. We describe the reasoner, how to write an inference rule and check which explicit knowledge was used to infer a new one. Real cases examples will be provided to show the capabilities of our reasoner and the associated language developed to express inference rules. We show how HORUS allows the user to understand the logical process over which each new RDF triple has been generated.