ISCA & IEEE Workshop on Spontaneous Speech Processing and Recognition

April 13-16, 2003
Tokyo Institute of Technology, Tokyo, Japan

The Map Task Corpus of Spoken Russian

Veronika Makarova (1), Valery A. Petrushin (2)

(1) Meikai University, National Institute of Advanced Industrial Science and Technology, Japan
(2) Accenture Technology Labs, Accenture, Chicago, USA

This paper describes the purposes, structure and applications as well as the speakers, material, recording, digitizing, labeling and storage procedures of Map Task Corpus of spoken Russian. The database is comprised of recordings of 116 spontaneous unscripted taskoriented dialogues produced by 64 native speakers of Russian while performing the task of marking a route on a printed map. The task was performed only via verbal communication, other forms (such as eye contact or gestures) being excluded in experimental settings. The total duration of the recorded dialogues is 18 hours. The database is constructed as a material source for theoretical and applied linguistics, language teaching, psycholinguistics, communication, speech processing, recognition as well as for interdisciplinary applications.


Full Paper

Bibliographic reference.  Makarova, Veronika / Petrushin, Valery A. (2003): "The map task corpus of spoken Russian", in SSPR-2003, paper TAP2.