An ontology based approach to reduce the cross-language translation ambiguity

In this paper we proposed an ontology-based approach to overcome the translation ambiguity problem of bilingual information retrieval. Based on the domain ontology, we can inference the most possible combination of translated words. Each combination of translated words has a probability, and we can eliminate unmeaningful translated words. A metric based on the ontology, called ontological chain, has been proposed for co-occurrence concept evaluation. Based on the ontology chain, we can calculate the best translation upshot of all possible terms. The experimental results show that the proposed approach can effectively remove the irrelevant results and improve the precision.