Papers › A Two-Stage Prediction-Aware Contrastive Learning Framework for Multi-Intent NLU

A Two-Stage Prediction-Aware Contrastive Learning Framework for Multi-Intent NLU

5 May 2024arXiv:2405.02925archive 2025-07-28

Guanhua Chen, Yutong Yao, Derek F. Wong, Lidia S. Chao

Multi-intent natural language understanding (NLU) presents a formidable challenge due to the model confusion arising from multiple intents within a single utterance. While previous works train the model contrastively to increase the margin between different multi-intent labels, they are less suited to the nuances of multi-intent NLU. They ignore the rich information between the shared intents, which is beneficial to constructing a better embedding space, especially in low-data scenarios. We introduce a two-stage Prediction-Aware Contrastive Learning (PACL) framework for multi-intent NLU to harness this valuable knowledge. Our approach capitalizes on shared intent information by integrating word-level pre-training and prediction-aware contrastive fine-tuning. We construct a pre-training dataset using a word-level data augmentation strategy. Subsequently, our framework dynamically assigns roles to instances during contrastive fine-tuning while introducing a prediction-aware contrastive loss to maximize the impact of contrastive learning. We present experimental results and empirical analysis conducted on three widely used datasets, demonstrating that our method surpasses the performance of three prominent baselines on both low-data and full-data scenarios.

PaperPDF

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

No code repository is listed for this paper in the archive or in Syntology's graph.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Contrastive LearningData AugmentationIntent DetectionNatural Language UnderstandingPredictionSemantic Frame ParsingSlot Filling

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Intent Detection MixATIS TFMN (PACL) Accuracy 82.9 #3 of 15 Archive leaderboard report
Intent Detection MixATIS SLIM (PACL) Accuracy 81.9 #4 of 15 Archive leaderboard report
Intent Detection MixATIS RoBERTa (PACL) Accuracy 79.1 #7 of 15 Archive leaderboard report
Intent Detection MixSNIPS TFMN (PACL) Accuracy 97.4 #6 of 16 Archive leaderboard report
Intent Detection MixSNIPS SLIM (PACL) Accuracy 96.9 #10 of 16 Archive leaderboard report
Intent Detection MixSNIPS RoBERTa (PACL) Accuracy 96.5 #13 of 16 Archive leaderboard report
Slot Filling MixATIS SLIM (PACL) Micro F1 87.3 #13 of 15 Archive leaderboard report
Slot Filling MixATIS TFMN (PACL) Micro F1 86.7 #14 of 15 Archive leaderboard report
Slot Filling MixATIS RoBERTa (PACL) Micro F1 86.0 #15 of 15 Archive leaderboard report
Slot Filling MixSNIPS SLIM (PACL) Micro F1 96.8 #2 of 16 Archive leaderboard report
Slot Filling MixSNIPS TFMN (PACL) Micro F1 96.3 #6 of 16 Archive leaderboard report
Slot Filling MixSNIPS RoBERTa (PACL) Micro F1 96.2 #7 of 16 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Contrastive Learning

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections