Abstract

The Semantic Web is an effort to establish standards and mechanisms that allow computers to reason more easily about the semantics of Web resources (documents, data etc.). Ontologies play a central role in this endeavour; given an OWL ontology for a knowledge domain, one can publish on the Web machine-readable data pertaining to that domain (e.g., catalogues of products, their features etc.), with the data having formally defined semantics based on the conceptualization of the ontology. Several OWL syntaxes have been developed, but people unfamiliar with formal knowledge representations often have difficulties understanding them.

This talk is organized in three parts, with the first part discussing natural language generation methods that allow end-users to view ontology-based knowledge representations in the form of automatically generated texts in multiple natural languages. Arguing against the pipeline architecture of most natural language generation systems, where the stages are greedily considered one after the other, the second part aims to further improve the quality of generated texts by introducing an Integer Linear Programming model that jointly considers content selection, lexicalization, sentence aggregation, and a limited form of referring expression generation. Lastly, in an attempt to reduce the tedious and time consuming manual authoring of domain-dependent linguistic resources that NLG needs to produce high quality texts, the last part examines text mining and machine learning methods to automatically or semi-automatically extract the most important of the resources from the Web.