| 1 |
Dynamic Grammars with Lookahead Composition for WFST-based Speech Recognition
|
|
|
|
In: http://www.gavo.t.u-tokyo.ac.jp/~mine/paper/PDF/2012/INTERSPEECH_1272_t2012-9.pdf
|
|
BASE
|
|
Show details
|
|
|
| 2 |
INTERSPEECH 2011 Painless WFST cascade construction for LVCSR- Transducersaurus
|
|
|
|
In: http://www.gavo.t.u-tokyo.ac.jp/~mine/paper/PDF/2011/INTERSPEECH_p1537-1540_t2011-8.pdf
|
|
BASE
|
|
Show details
|
|
|
| 3 |
Open-Source WFST tools for LVCSR Cascade Construction
|
|
|
|
In: http://www.gavo.t.u-tokyo.ac.jp/~mine/paper/PDF/2011/FSMNLP_t2011-7.pdf
|
|
BASE
|
|
Show details
|
|
|
| 4 |
Open Source WFST tools for LVCSR cascade development
|
|
|
|
In: http://www.aclweb.org/anthology-new/W/W11/W11-4409.pdf
|
|
BASE
|
|
Show details
|
|
|
| 5 |
Failure transitions for Joint n-gram Models and G2P Conversion
|
|
|
|
In: http://www.gavo.t.u-tokyo.ac.jp/~mine/paper/PDF/2013/INTERSPEECH_p1821-1825_t2013-8.PDF
|
|
BASE
|
|
Show details
|
|
|
| 6 |
CLEF 2009 Question Answering Experiments at Tokyo Institute of Technology
|
|
|
|
In: http://ceur-ws.org/Vol-1175/CLEF2009wn-QACLEF-HeieEt2009.pdf
|
|
Abstract:
In this paper we describe the experiments carried out at Tokyo Institute of Technology for the CLEF 2009 Question Answering on Speech Transcriptions (QAST) task, where we participated in the English track. We apply a non-linguistic, data-driven approach to Question Answering (QA). Relevant sentences are rst retrieved from the supplied corpus, using a language model based sentence retrieval module. Our probabilistic answer extraction module then pinpoints exact answers in these sentences. In this year's QAST task the question set contains both factoid and non-factoid questions, where the non-factoid questions ask for denitions of given named entities. We do not make any adjustments of our factoid QA system to account for non-factoid questions. Moreover, we are presented with the challenge of searching for the right answer in a relatively small corpus. Our system is built to take advantage of redundant information in large corpora, however, in this task such redundancy is not available. The results show that our QA framework does not perform well on this task: we end last of four participating teams in seven out of eight runs. However, our performance does not regress as automatic transcriptions of speeches or questions are used instead of manual transcriptions. Thus the only run in which we are not placed last, is the most dicult task, where spoken questions and ASR transcriptions with high WER are used.
|
|
Keyword:
Categories and Subject Descriptors H.3 [Information Storage and Retrieval; Experimentation Keywords Question answering; H.3.3 Information Search and Retrieval; H.3.4 Systems and Software General Terms Measurement; Performance; Questions beyond factoids
|
|
URL: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.664.1724
http://ceur-ws.org/Vol-1175/CLEF2009wn-QACLEF-HeieEt2009.pdf
|
|
BASE
|
|
Hide details
|
|
|
|
|
