AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Multi-Task Deep Learning for User Intention Understanding in Speech Interaction Systems

Yishuang Ning, Jia Jia, Zhiyong Wu, Runnan Li, Yongsheng An, Yanfeng Wang, Helen Meng

AAAI 2017 · Label-efficient and robust learning › Recognition and long-tail · approach: Multilingual language representation pretraining · 18 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add Bayesian network on top of LSTM outputs (ablation)LSTM · Sogou intention prominenceF1: 0.797 -> 0.866Improved
use LSTM instead of CRFintention prominence model · Sogou focus detectionF1: 0.739 -> 0.759Improved
use LSTM instead of CRFemphasis detection · Sogou emphasis detectionF1: 0.066 -> 0.575Improved
add emphasis by speech to focus by text (ablation)LSTM+BN · Sogou intention prominenceF1: +2.8%Improved
add location information to focus and emphasis (ablation)LSTM+BN · Sogou intention prominenceF1: +3%Improved
multi-task learning for focus and emphasis detection (ablation)LSTM · Sogou focus and emphasis detectionF1Improved
add more unlabeled data for SAE pretraining (ablation)LSTM+BN · Sogou intention prominenceF1Improved
feed intention prominence rather than original utterances into searchSogou search · 32 utterancestop-10 coverage ratio: 65.25% -> 72.25%Improved
use LSTM+BN instead of BNintention prominence model · CPU training timeCPU time (s): 6.8 -> 1.7520 x10^4Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy