Multi-Label Few-Shot Learning for Aspect Category Detection
This work addresses a domain-specific problem in natural language processing for sentiment analysis by extending few-shot learning to multi-label scenarios, which is incremental but addresses a known bottleneck in existing methods.
The paper tackles the problem of aspect category detection in sentiment analysis under few-shot learning conditions, where sentences can contain multiple aspect categories, and proposes a multi-label few-shot learning method based on prototypical networks with attention mechanisms and dynamic thresholding, achieving significant performance improvements over baselines on three datasets.
Aspect category detection (ACD) in sentiment analysis aims to identify the aspect categories mentioned in a sentence. In this paper, we formulate ACD in the few-shot learning scenario. However, existing few-shot learning approaches mainly focus on single-label predictions. These methods can not work well for the ACD task since a sentence may contain multiple aspect categories. Therefore, we propose a multi-label few-shot learning method based on the prototypical network. To alleviate the noise, we design two effective attention mechanisms. The support-set attention aims to extract better prototypes by removing irrelevant aspects. The query-set attention computes multiple prototype-specific representations for each query instance, which are then used to compute accurate distances with the corresponding prototypes. To achieve multi-label inference, we further learn a dynamic threshold per instance by a policy network. Extensive experimental results on three datasets demonstrate that the proposed method significantly outperforms strong baselines.