Towards a Conceptual Search for Vietnamese Legal Text

In this paper, a system of search engine is built based on documents of Vietnamese legal system. The key factor of this system is the ability of indexing documents in several aspects: not only on the whole text but also on logical structures and ontologies of Vietnamese legal documents. There are two important phases in the system; firstly, focusing on the recognition of the structure of Vietnamese legal text and the extraction of ontology of documents stored in the database of Vietnam Ministry of Justice; secondly, building an automated information retrieval system for the advanced search on demand of Vietnamese citizens. We study Vietnamese legal domain in both linguistic features and patterns of recognition. Experimental result on recognizing logical structures got of 64.37% on assumption annotation, 64.15% on provision annotation and 75.76% on sanction annotation in the

F

β

 = 1

score on the pre-built corpus of Vietnamese Enterprise Law articles. We also evaluated the precision of classifying ontology concepts and got 86.4 % in result on testing sample set of Vietnamese legal ontology. The goal is that a search engine is proposed with indexed logical and semantic properties based on works of the first phase.