Title :
Turkish LVCSR: towards better speech recognition for agglutinative languages
Author :
Çarki, Kenan ; Geutner, Petra ; Schultz, Tanja
Author_Institution :
Interactive Syst. Labs., Karlsruhe Univ., Germany
Abstract :
The Turkish language belongs to the Turkic family. All members of this family are close to one another in terms of linguistic structure. Typological similarities are vowel harmony, verb-final word order and agglutinative morphology. This latter property causes a very fast vocabulary growth resulting in a large number of out-of-vocabulary words. In this paper we describe our first experiments in a speaker independent LVCSR engine for Modern Standard Turkish. First results on our Turkish speech recognition system are presented. The currently best system shows very promising results achieving 16.9% word error rate. To overcome the OOV-problem we propose a morphem-based and the Hypothesis Driven Lexical Adaptation approach. The final Turkish system is integrated into the multilingual recognition engine of the GlobalPhone project
Keywords :
linguistics; natural languages; speech recognition; GlobalPhone project; Hypothesis Driven Lexical Adaptation approach; Modern Standard Turkish; OOV-problem; Turkic family; Turkish LVCSR; Turkish language; agglutinative languages; agglutinative morphology; linguistic structure; morphem-based approach; multilingual recognition engine; out-of-vocabulary words; speaker independent LVCSR engine; speech recognition; typological similarities; verb-final word order; very fast vocabulary growth; vowel harmony; word error rate; Databases; Engines; Interactive systems; Laboratories; Morphology; Natural languages; Speech analysis; Speech recognition; Testing; Vocabulary;
Conference_Titel :
Acoustics, Speech, and Signal Processing, 2000. ICASSP '00. Proceedings. 2000 IEEE International Conference on
Conference_Location :
Istanbul
Print_ISBN :
0-7803-6293-4
DOI :
10.1109/ICASSP.2000.861971
