Linguistic ontologies, most notably WordNet [1], have been shown to be a valuable resource for a variety of natural language processing applications. Presently, linguistic ontologies are largely constructed by hand, which is both difficult and expensive. A central problem that demands an automated solution is the discovery and incorporation of lexical semantic relations, or semantic relations b...

We investigate from the competence standpoint two recent models of lexical semantics, algebraic conceptual representations and continuous vector models. Characterizing what it means for a speaker to be competent in lexical semantics remains perhaps the most significant stumbling block in reconciling the two main threads of semantics, Chomsky’s cognitivism and Montague’s formalism. As Partee (19...

Automatic extraction of semantic information, if successful, offers to languages with little or poor resources, the prospects of creating ontological resources inexpensively, thus providing support for common-sense reasoning applications in those languages. In this paper we explore the automatic extraction of synonymy information from large corpora using two complementary techniques: a generic ...

Automatic extraction of semantic information, if successful, offers to languages with little or poor resources, the prospects of creating ontological resources inexpensively, thus providing support for commonsense reasoning applications in those languages. In this paper we explore the automatic extraction of synonymy information from large corpora using two complementary techniques: a generic b...

Semantic dictionaries are widely used in natural language processing studies. Broad range of methods have been proposed to construct semantic dictionaries until this time. In this study, a brief summary for semantic relation pair extraction is introduced. An overview of selected approaches is given and results are compared to each other for hyponymy, holonymy, and antonymy relations.