DFKI-LT - Multilingual Terminology Acquisition for Ontology-based Information Extraction
Multilingual Terminology Acquisition for Ontology-based Information Extraction
4 Proceedings of the 10th Terminology and Knowledge Engineering Conference,
We present current work on the automated acquisition of multilingual terms for labels of ontologies in the financial domain. The main approach consists in harvesting multilingual web pages of stock exchanges, and to extract the relevant data encoded in HTML feature structures from them. Out of these feature structures, we extract and align the multilingual vocabulary that can be used either in labels of classes or properties defined in ontologies, or as part of the value of properties. We also discuss the use of standardized terminological frameworks for improving and validating the results of the automated extraction of multilingual term candidates.
Files: BibTeX, Paper 12 pp166-176.pdf