@inproceedings{yu-liu-2026-kevinyu66,
    title = "kevinyu66 at {S}em{E}val-2026 Task 3: A Retrieval-Augmented {LLM} System for Aspect{--}Opinion Triplet Extraction",
    author = "Yu, Kuanlin  and
      Liu, Wen-Ni",
    editor = "Kochmar, Ekaterina  and
      Ghosh, Debanjan  and
      North, Kai  and
      Komachi, Mamoru",
    booktitle = "Proceedings of the 20th {I}nternational {W}orkshop on {S}emantic {E}valuation (2026)",
    month = jul,
    year = "2026",
    address = "San Diego, California, USA",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl-workshops/2026.semeval-1.16/",
    pages = "108--114",
    ISBN = "979-8-89176-414-9",
    abstract = "This paper describes our system used in the SemEval-2026 Task 3: Dimensional Aspect-Based Sentiment Analysis. To address the inherent subjectivity and nuanced emotional expressions in this task, we propose a Retrieval-Augmented Generation (RAG) framework based on Large Language Models (LLMs) for sentiment triplet extraction. Our approach leverages a dynamic retrieval mechanism to identify semantically similar training examples, which are then integrated into the prompts as in-context demonstrations. This strategy effectively guides the model{'}s inference process by providing relevant linguistic patterns and emotional contexts. Our implementation is available at https://github.com/Kevinyu66/dimaste."
}