Quantcast

Semantic classification-based approach for the self-enrichment of LMF normalized dictionaries

Research paper by Imen Elleuch, Bilel Gargouri, Abdelmajid Ben Hamadou

Indexed on: 23 Oct '16Published on: 22 Oct '16Published in: Lingua



Abstract

In this paper, we propose an approach for the self-enrichment of LMF normalized dictionaries with semantic classes. This approach is composed of four processes. The first one deals with the semantic classification process based on the proposition of Gaston Gross. The second benefits from the available subject fields linked to the meanings of the dictionary lexical entries in order to identify primary semantic classes. The third process refines the subject-field-based enrichment by a specific analysis of the textual content of the lexical entries’ meanings. The final process consists in carrying out a complete syntactico-semantic analysis on the textual content related to the meanings of the lexical entry. In addition, the third and the fourth processes optimize the enrichment process by attributing the same semantic classes to the synonymous meanings. The proposed approach was consolidated by the implementation of an enrichment system which was tested on an available normalized Arabic dictionary.

Figure 10.1016/j.lingua.2016.09.001.0.jpg
Figure 10.1016/j.lingua.2016.09.001.1.jpg
Figure 10.1016/j.lingua.2016.09.001.2.jpg
Figure 10.1016/j.lingua.2016.09.001.3.jpg
Figure 10.1016/j.lingua.2016.09.001.4.jpg
Figure 10.1016/j.lingua.2016.09.001.5.jpg
Figure 10.1016/j.lingua.2016.09.001.6.jpg