Interventional Aspect-Based Sentiment Analysis
This work addresses robustness issues in sentiment analysis for applications like review analysis, but it is incremental as it builds on existing neural-based approaches with a causal adjustment.
The paper tackled the problem of poor robustness in aspect-based sentiment analysis when encountering confounders like non-target aspects, and the result was that their proposed method improved performance on the Aspect Robustness Test Set while maintaining accuracy on the original test set.
Recent neural-based aspect-based sentiment analysis approaches, though achieving promising improvement on benchmark datasets, have reported suffering from poor robustness when encountering confounder such as non-target aspects. In this paper, we take a causal view to addressing this issue. We propose a simple yet effective method, namely, Sentiment Adjustment (SENTA), by applying a backdoor adjustment to disentangle those confounding factors. Experimental results on the Aspect Robustness Test Set (ARTS) dataset demonstrate that our approach improves the performance while maintaining accuracy in the original test set.