Volltext-Downloads (blau) und Frontdoor-Views (grau)

Von der Tonbandaufnahme zur integrierten Text-Ton-Datenbank. Instrumente für die Arbeit mit Gesprächskorpora

  • The development of tools for computer-assisted transcription and analysis of extensive speech corpora is one main issue at the Institute of German Language (IDS) and the Institute of Natural Language Processing (IMS). Corpora of natural spoken dialogue have been transcribed, and the analogue recordings of these discourses are digitized. An automatic segmentation system is employed which is based on Hidden Markov Models. The orthographic representation of the speech signal is transformed into a phonetic representation, the phonetic transcription is transformed into a system-inter­nal representation, and the time alignment between text and speech signal follows. In this article, we also describe the retrieval software Cosmas II and its special features for searching discourse transcripts and playing time aligned passages.

Export metadata

Additional Services

Share in Twitter Search Google Scholar


Author:Franck Bodmer Mory, Marcus L. Fach, Rudolf Schmidt, Wilfried Schütte
Parent Title (German):Romanistische Korpuslingustik: Korpora und gesprochene Sprache
Series (Serial Number):ScriptOralia (126)
Place of publication:Tübingen
Editor:Claus D. Pusch, Wolfgang Raible
Document Type:Part of a Book
Year of first Publication:2002
Date of Publication (online):2015/11/02
GND Keyword:Automatische Spracherkennung; Konversationsanalyse; Korpus <Linguistik>; Transkription
First Page:209
Last Page:243
Dewey Decimal Classification:400 Sprache / 410 Linguistik
Open Access?:ja
BDSL-Classification:Sprache im 20. Jahrhundert. Gegenwartssprache
Leibniz-Classification:Sprache, Linguistik
Linguistics-Classification:Gesprächsforschung / Gesprochene Sprache
Licence (German):Es gilt das UrhG