@inproceedings{wu-etal-2026-id,
    title = "From {ID} to {LLM}: Rethinking Representation Learning for Recommendation",
    author = "Wu, Song-Li  and
      Du, Zhaocheng  and
      Gan, Weinan  and
      Wang, Jingyi  and
      Wang, Xianquan",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.acl-long.179/",
    pages = "3893--3905",
    ISBN = "979-8-89176-390-6",
    abstract = "Recent studies indicate a fundamental incompatibility between ID representations and language model (LM) representations, as they capture behavioral and semantic spaces respectively. This mismatch leads LM representations to consistently underperform ID representations in recommendation tasks. In this work, we revisit this problem and show, from an information-theoretic perspective, that LLM representations retain all discriminative information in ID representations. Based on this, we introduce a Profile-then-Embedding (PtE) framework for recommendation, consisting of a Profile Stage, in which semantic user and item profiles are generated jointly through LLM-based bidirectional reasoning over user-item interactions, and a Personalized Embedding Stage, which encodes these profiles into task-aligned recommendation embeddings. We demonstrate PtE{'}s effectiveness across three benchmark datasets, including cold-start and long-tail scenarios, achieving substantial gains in both discriminative and generative recommendation models."
}