NLP@AIIA 2008
This workshop aims at exploring new trends of research and development on the integration between ontologies and linguistic resources, where complementary levels of semantic specification meet.
In general, the actual practice of enhancing ontologies with linguistic content basically depends on the multifariousness of lexical resources and on the explicit linguistic information they expose. Multilingual scenarios also demand for a proper lexicalization of ontological content according to different idioms and languages. From simple vocabularies of terms to wordnet-like structures, distinct lexical models need a solid and comprehensive framework of representation to enable a full-operational integration with ontologies. The integration between frame-based lexical databases and ontologies complicates the overall scenario and constitutes another important aspect of the above-mentioned integration process and a relatively brand-new trend in the scientific community. These topics have recently become more and more important also in the challenge of combining Human-Language Technologies with Semantic Web technologies, towards the creation of next-generation knowledge systems.
The above-mentioned integration process affects ontology learning too. Different automatic methodologies and techniques are employed to build ontologies from variable-scaled linguistic resources such as annotated corpora, unstructured texts, terminologies, lexical databases, etc. Although it is important to adopt automatic techniques and tools to ‘skim off’ knowledge from heterogeneous data sources, the essential limit of ontology learning lays in the quality of the obtained resources: in order to make extracted ontologies robust and consistent with cognitive models of knowledge, human-based conceptual analysis and possible re-organization of the overall ontological structure are needed, together with suitable comparisons with pre-existing (domain and foundational) ontologies.
By means of this workshop the NLP group of the Italian Association for Artificial Intelligence aims at gathering together experts from the scientific community to present and discuss original approaches in the above mentioned scenario.
