SuperLectures.com

A SEGMENT-LEVEL CONFIDENCE MEASURE FOR SPOKEN DOCUMENT RETRIEVAL

Full Paper at IEEE Xplore

Spoken Document Processing

Přednášející: Gregory Senay, Autoři: Gregory Senay, Georges Linarès, Benjamin Lecouteux, University of Avignon, France

This paper presents a semantic confidence measure that aims to predict the relevance of automatic transcripts for a task of Spoken Document Retrieval. The proposed predicting method relies on the combination of Automatic speech Recognition (ASR) confidence measure and a local semantic compacity index (SCI), that estimates the relevance of the words regarding the semantic context in which they occurred. Experiments are conducted on the French Broadcast news corpus ESTER, by simulating a classical SDR usage scenario : users submit text-queries to a search engine that is expected to return the most relevant documents regarding the query. Results demonstrate the interest of using semantic level information to predict the transcription indexability.


  Přepis řeči

|

  Slajdy

Zvětšit slajd | Zobrazit všechny slajdy

0:00:16

  1. slajd

0:01:02

  2. slajd

0:01:39

  3. slajd

0:01:59

  4. slajd

0:02:16

  5. slajd

0:03:15

  6. slajd

0:03:21

  7. slajd

0:04:08

  8. slajd

0:05:04

  9. slajd

0:05:34

 10. slajd

0:06:18

 11. slajd

0:07:20

 12. slajd

0:08:04

 13. slajd

0:09:16

 14. slajd

0:11:20

 15. slajd

  Komentáře

Please sign in to post your comment!

  Informace o přednášce

Nahráno: 2011-05-27 14:45 - 15:05, Panorama
Přidáno: 9. 6. 2011 02:17
Počet zhlédnutí: 24
Rozlišení videa: 1024x576 px, 512x288 px
Délka videa: 0:15:46
Audio stopa: MP3 [5.38 MB], 0:15:46