Where to Go Next for Recommender Systems? ID- vs. Modality-based Recommender Models Revisited
Open accessAbstract
Recommendation models that utilize unique identities (IDs for short) to represent distinct users and items have been state-of-the-art (SOTA) and dominated the recommender systems (RS) literature for over a decade. Meanwhile, the pre-trained modality encoders, such as BERT [9] and Vision Transformer [11], have become increasingly powerful in modeling the raw modality features of an item, such as text and images. Given this, a natural question arises: can a purely modality-based recommendation model (MoRec) outperforms or matches a pure ID-based model (IDRec) by replacing the itemID embedding with a SOTA modality encoder? In fact, this question was answered ten years ago when IDRec beats MoRec by a strong margin in both recommendation accuracy and efficiency.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Using only textual features as inputs to recommendation models, leveraging pre-trained language models, can alleviate cold-start problems and enable cross-domain recommendations.
“Additionally, some recent works [114-117] have explored using only textual features as inputs to recommendation models, leveraging pre-trained language models to alleviate coldstart problems and enable cross-domain recommendations.”
From When large language models meet personalization: perspectives of challenges and opportunities (2024) · cites this paper as [117]
LLM-Embedding-Based Recommenders use embeddings from LLMs in traditional systems to capture user preferences.
“LLM-Embedding-Based Recommenders use embeddings from LLMs in traditional systems to capture user preferences (Yuan et al., 2023; Xi et al., 2024a).”
From Decoding in Latent Spaces for Efficient Inference in LLM-based Recommendation (2025) · cites this paper as Yuan et al., 2023
The citation lists haven’t been fetched yet
References 47 · Cited by 200
- Authors
- Yuan, Z., Yuan, F., Song, Y., Li, Y., Fu, J., Yang, F., Pan, Y., & Ni, Y.
- Affiliations
- Westlake University (CN)
- Zhejiang Lab (CN)
- Venue
- Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval
- Publisher
- ACM
- Published
- 2023-07-18
- Pages
- 2639-2649
- OpenAlex
- W4384648324
- Cited by
- 200
- References
- 47
- Citations by year
- 2026: 61 · 2025: 72 · 2024: 59 · 2023: 8
- FWCI
- 51.44
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR) (publishedVersion, cc-by)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- the Research Center for Industries of the Future (No.WU2022C030)
- the Key Research Project of Zhejiang Lab (No.2022PG0AC02)
- Topics
- Physical Sciences › Computer Science › Information Systems › Recommender Systems and Techniques
- Physical Sciences › Computer Science › Information Systems › Information Retrieval and Search Behavior
- Social Sciences › Decision Sciences › Management Science and Operations Research › Advanced Bandit Algorithms Research
- Keywords
- recommender systems, item embeddings, BERT, vision transformer, user and item representations, recommendation efficiency
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
