There exists a vast amount of lexicographical data, including dictionaries, thesauri, Wordnet and so on, that are distinctly providing various information about words. The automatic alignment of such resources is currently a challenging task.
In a recent attempt to address the task of monolingual word sense alignment (MWSA), we created a set of monolingual datasets containing manual annotations of semantic relationships between senses across resources. However, the sheer volume and heterogeneity of data make human annotation limited. Instead, we would like to have a computer annotate all data with the structure of our interest. Normally, we are interested in relations between word definitions, a.k.a. senses.
Information Extraction (IE) automatically finds relevant entities or relations (including facts and events) in natural language texts. More specifically, the task of Relation Extraction (RE) aims to recognize and extract instances of semantic relations between entities or concepts mentioned in these texts. Many applications in information extraction, natural language understanding, information retrieval require an understanding of the semantic relations between entities. The main objective of capturing structured relational knowledge about lexical terms has been the motivating force underlying many projects in lexical acquisition, information extraction, and the construction of semantic taxonomies.
What has been done so far?