Result Details
Extracting Situation Frames from non-English Speech: Evaluation Framework and Pilot Results
Glembek Ondřej, Ing., Ph.D., DCGM (FIT)
Narayanan Shrikanth, prof., PhD
This paper describes the first evaluation framework for the extractionof Situation Frames - structures describing humanitarianassistance needs - from non-English speech audio, conductedfor the DARPA LORELEI (Low Resource Languagesfor Emergent Incidents) program. Participants in LORELEI hadto process audio from a variety of sources, in non-English languages,and extract the information required to populate SituationFrames describing whether any need is mentioned, thetype of need present and where the need exists. The evaluationwas conducted over a period of 10 days and attracted submissionsfrom 6 teams, each team spanning multiple organizations.Performance was evaluated using precision-recall curves. Theresults are encouraging, with most teams showing some capabilityto detect the type of situation discussed, but more workwill be required to connect needs to specific locations.
speech recognition, speech analysis, performanceevaluation, natural language processing
@inproceedings{BUT163392,
author="Nikolaos {Malandrakis} and Ondřej {Glembek} and Shrikanth {Narayanan}",
title="Extracting Situation Frames from non-English Speech: Evaluation Framework and Pilot Results",
booktitle="Proceedings of Interspeech 2017",
year="2017",
journal="Proceedings of Interspeech",
volume="2017",
number="8",
pages="2123--2127",
publisher="International Speech Communication Association",
address="Stockholm",
doi="10.21437/Interspeech.2017-226",
issn="1990-9772",
url="https://www.semanticscholar.org/paper/Extracting-Situation-Frames-from-Non-English-and-Malandrakis-Glembek/047ebaf46d648fea9e5c2c64873c629f7fd53a00"
}