An Intelligent Visual Dictionary for Italian Sign Language

Sign languages are visual-gestural languages developed mainly in deaf communities; their tempo-spatial nature makes it difficult to write them, yet several transcription systems are available for them. Most sign language dictionaries interact with users via a transcription-based interface; thus users need to be expert of that specific transcription system. The e-LIS dictionary is the first web bidirectional dictionary for Italian sign language-Italian; using the current e-LIS interface, users can define a sign interacting with intuitive iconic images, ignoring the underlying transcription system. Nevertheless this interface assumes that its users are expert signers, knowledgeable about the formational rules of signs. The e-LIS ontology, which specifies how to form a sign, allows even non-expert signers to use the dictionary from Italian sign language to Italian. This is realised through a novel visual interface for transparently browsing and querying the e-LIS ontology and the underlying database. The interface prototype was designed following the user centred design methodology. As such, our work constitutes the first attempt at making the e-LIS dictionary an intelligent visual dictionary, usable by learners of Italian sign language who are not expert signers yet. This paper reports on the design of the first prototype of the ontology-based visual interface, and outlines its evaluation plan.