Title :
Assessing search term strength in spoken term detection
Author :
Torbati, Amir Hossein Harati Nejad ; Picone, J.
Author_Institution :
Dept. of Electr. & Comput. Eng., Temple Univ., Philadelphia, PA, USA
Abstract :
Spoken term detection (STD) is an extension of text-based searching that allows users to type keywords and search audio files containing recordings of spoken language. Performance is dependent on many external factors such as the acoustic channel, the language and the confusability of the search term. Unlike text-based searches, the quality of the search term plays a significant role in the overall perception of the usability of the system. In this paper, we present a system that predicts the strength of a search term from its spelling that is based on an analysis of spoken term detection output from several spoken term detection systems that participated in the NIST 2006 STD evaluation. We show that approximately 57% of the correlation can be explained from the search term, but that a significant amount of the confusability is due to other acoustic modeling issues.
Keywords :
acoustic signal processing; information retrieval; speaker recognition; text analysis; NIST 2006 STD evaluation; acoustic channel; acoustic modeling; audio file search; correlation; keyword; search term confusability; search term strength assessment; spelling; spoken language recording; spoken term detection; system usability; text-based searching; Acoustics; Conferences; Error analysis; Keyword search; NIST; Speech; Speech recognition; information retrieval; spoken term detection; voice keyword search;
Conference_Titel :
Cognitive Methods in Situation Awareness and Decision Support (CogSIMA), 2013 IEEE International Multi-Disciplinary Conference on
Conference_Location :
San Diego, CA
Print_ISBN :
978-1-4673-2437-3
DOI :
10.1109/CogSIMA.2013.6523832