Search papers, labs, and topics across Lattice.
This paper introduces a Zero-Initialized Residual Complex Projection (ZRCP) framework for Aspect-Based Sentiment Analysis (ABSA) that projects textual features into a complex semantic space, using phase to disentangle sentiment polarities and amplitude for semantic intensity. They also propose an Anti-collision Masked Angle Loss to mitigate false-negative collisions in contrastive learning, particularly for high-frequency aspects. Experiments show that ZRCP achieves state-of-the-art results, with a Macro-F1 score of 0.8851, and geometric analyses confirm the importance of unconstrained amplitude in the complex space for robust sentiment disentanglement.
By projecting text into a complex semantic space, this model achieves SOTA on aspect-based sentiment analysis by disentangling sentiment polarity from semantic intensity.
Aspect-Based Sentiment Analysis (ABSA) is fundamentally challenged by representation entanglement, where aspect semantics and sentiment polarities are often conflated in real-valued embedding spaces. Furthermore, standard contrastive learning suffers from false-negative collisions, severely degrading performance on high-frequency aspects. In this paper, we propose a novel framework featuring a Zero-Initialized Residual Complex Projection (ZRCP) and an Anti-collision Masked Angle Loss,inspired by quantum projection and entanglement ideas. Our approach projects textual features into a complex semantic space, systematically utilizing the phase to disentangle sentiment polarities while allowing the amplitude to encode the semantic intensity and lexical richness of subjective descriptions. To tackle the collision bottleneck, we introduce an anti-collision mask that elegantly preserves intra-polarity aspect cohesion while expanding the inter-polarity discriminative margin by over 50%. Experimental results demonstrate that our framework achieves a state-of-the-art Macro-F1 score of 0.8851. Deep geometric analyses further reveal that explicitly penalizing the complex amplitude catastrophically over-regularizes subjective representations, proving that our unconstrained-amplitude and phase-driven objective is crucial for robust, fine-grained sentiment disentanglement.
