NatS Oberseminar 23. June 2005

Comparison of matching algorithms for Example-based Machine Translation

Vanesa Espin Martin

In my talk I willgive first an overview of Example-Based Machine Translation (EBMT), what is, what EBMT systems are, its architecture and its operation keys (retrieval and adaptation).

Further I will explain how do I prepared my Corpus (Spanish-English Paralell Corpus about Tourism in Spain) to face EBMT.

The last part of the talk will consider: Information Retrieval from the Corpus: matching methods for retrieving translation examples similar to a given input (Edit Distance, String Matching, Word by Word with the help of a Thesaurus) and some comparisons between its efficiencies: exact matches, similarity, fragments matching,etc.)"