Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces the SeqLab framework, which enhances cross-lingual aspect-based sentiment analysis (ABSA) by integrating a sequence-to-sequence model with an auxiliary sequence-labelling task to improve aspect term recognition and sentiment predictions. By employing aspect-code switching (ACS), the authors generate additional training data through translation-based swapping of aspect terms, significantly improving the model's cross-lingual capabilities. Evaluated across eleven languages and multiple domains, the proposed method achieves state-of-the-art results in the end-to-end ABSA task, particularly excelling in the complex target-aspect-sentiment detection (TASD) task.
Aspect-code switching boosts cross-lingual sentiment analysis performance, achieving state-of-the-art results across multiple languages and domains.
Cross-lingual aspect-based sentiment analysis (ABSA) transfers knowledge from a source language with annotated data to a target language, enabling fine-grained sentiment analysis without annotated target-language data. While monolingual ABSA has seen significant progress, cross-lingual ABSA remains underexplored, especially for complex tasks involving multiple sentiment elements like target-aspect-sentiment detection (TASD). In this paper, we propose a novel SeqLab framework that enhances cross-lingual ABSA using a sequence-to-sequence model with an auxiliary sequence-labelling task performed by the encoder, enhancing aspect term recognition and sentiment predictions. Additionally, we incorporate aspect-code switching (ACS), a translation-based technique that swaps aspect terms between source and translated sentences, generating additional training data to enhance the model's cross-lingual understanding. We evaluate our approach across eleven languages, three domains, and two backbone models, surpassing previous state-of-the-art results for the commonly studied E2E-ABSA task. Unlike most prior work that relies solely on English as the source language, we systematically assess different source-target language pairs and extend our evaluation to the more challenging, yet underexplored TASD task in cross-lingual settings. Finally, we provide a detailed error analysis highlighting key challenges and limitations.