Lexico-Semantic Networks

This report will study various strategies of evaluation which are applied to ontologies and then develop approaches towards testing Lexico-semantic networks. We can get to know about lack of Lexico-semantic networks like Word Net by this report and it also develops a case for that evaluations. This report will provide description about Lexico-semantic networks, explains successful methods in Machine Translation evaluation and also mentions the principles of evaluation.

Lexico-semantic networks like Word Net etc are burst in to prominence because applications which are targeted for the web will now aim to improve the semantic dimensions of their performance. Best example for such application is, a lexical resource will help in retrieving information by providing easy query keyword disambiguation and improves the quality of retrieved search results. Because now huge no. of documents is available through the web, this resulted in the need for further sophistication. Other examples are, generating contents for specific documents like tourist phrasebooks automatically, sensing the emotions in text automatically.

The Lexical resources are developed by the various research groups and can help, generate, review such contents. They are not just simple dictionaries also contains huge amount of semantic contents. These networks also reserves common sense concepts, arrange them ontologically and describe the real-world via lexical knowledge. Now such resources are increasing in applications involving language technology not only in English. Now Word Net project is used by most of major languages and Word net not covers efforts like Concept Net attempt.

Now production of such networks and their application in diverse areas is increasing which leads to using evaluation methods in describing the quality of rival networks also to check whether they fulfill the expectations and increase their performance in applications. We can investigate the evaluation strategies for Lexico-semantic networks by using this report.

🔥2 Views