Sign languages are visual languages used in deaf communities, mainly. They simultaneously combine shapes, orientations and movements of the hands, as well as non-manual components, e.g., facial expressions. Their spatial nature makes it difficult to write or even transcribe them. Moreover there is not a one-to-one relation between a sign language and the related verbal language; a word may be represented by more than one sign; likewise, one sign may be translated into more than one word. For instance, there is a single sign in Italian for the "cut with scissors" expression (see 383.1 in [15]). This makes the creation of a sign language dictionary quite a challenging one.The e-LIS project, for which we worked, aims at the creation of the first Electronic bidirectional dictionary for Italian Sign Language -Italian (Verbal Language). The creation of the e-LIS dictionary requires a combination of expertise and skills from various fields. In this thesis, we focus on the creation of an ontology for e-LIS, here conceived as a means for analysing, representing and reasoning with the entities and relations of the sign language dictionary. The enriched knowledge provided by the e-LIS ontology will guide users when they look up in the e-LIS dictionary for a specific sign; in particular users who are not familiar with Italian sign language or the transcription systems should benefit from the expert navigation provided by our ontology. Moreover, our ontology can be integrated with the SEWASIE query tool, which allows for the automated extraction of signs from the dictionary database.This thesis concludes with several open questions at the intersection of knowledge representation and reasoning, sign and computational linguistics, human computer interaction, data base design, internet technologies, which most elicited our research interests.