Presentation	2004/12/15 Prediction Using Similarity based on TF・AoI Yasutomo KIMURA, Kenji ARAKI,
PDF Download Page	PDF download Page Link
Abstract(in Japanese)	(See Japanese page)
Abstract(in English)	We propose the prediction method using similarity based on TF^*AoI. In spoken dialogue, we generally use background, context and limited domain for prediction. Although n-gram is words dictation, we hope the prediction for utterance. Therefore we decide the prediction which is next utterance of the highest similarity utterance in a learning corpus. Our similarity is calculated by using weight based on TF・AoI (Term Frequency × Amount of Information). We confirmed 5 points improvement compared to the matching rate measure.
Keyword(in Japanese)	(See Japanese page)
Keyword(in English)	Prediction / Similarity / TF^*AoI
Paper #	NLC2004-91,SP2004-131
Date of Issue

Conference Information
Committee	NLC
Conference Date	2004/12/15(1days)
Place (in Japanese)	(See Japanese page)
Place (in English)
Topics (in Japanese)	(See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To	Natural Language Understanding and Models of Communication (NLC)
Language	JPN
Title (in Japanese)	(See Japanese page)
Sub Title (in Japanese)	(See Japanese page)
Title (in English)	Prediction Using Similarity based on TF・AoI
Sub Title (in English)
Keyword(1)	Prediction
Keyword(2)	Similarity
Keyword(3)	TF^*AoI
1st Author's Name	Yasutomo KIMURA
1st Author's Affiliation	Graduate School of Information Science and Technology, Hokkaido University()
2nd Author's Name	Kenji ARAKI
2nd Author's Affiliation	Graduate School of Information Science and Technology, Hokkaido University
Date	2004/12/15
Paper #	NLC2004-91,SP2004-131
Volume (vol)	vol.104
Number (no)	540
Page	pp.pp.-
#Pages	6
Date of Issue