Browse State-of-the-Art › Recommendation Systems

Recommendation Systems

1,997 papers with code · 55 benchmarks · 56 datasets archive 2025-07-28

GraphsKnowledge BaseMiscellaneous
Recommendation System in AI Research

A Recommendation System is a specialized AI-driven model that analyzes user preferences and behaviors to suggest relevant content, products, or services. It is widely used in domains like e-commerce, streaming platforms, social media, and personalized learning.

AI research in recommendation systems focuses on:
- Collaborative Filtering: Predicting user preferences based on similar users' choices.
- Content-Based Filtering: Recommending items based on user history and item characteristics.
- Hybrid Models: Combining multiple techniques for better accuracy.
- Deep Learning & Transformers: Using neural networks and self-attention mechanisms for personalized recommendations.
- Graph-Based Approaches: Leveraging knowledge graphs for relationship-aware recommendations.

Key challenges include data sparsity, scalability, and bias mitigation. Cutting-edge research explores reinforcement learning, explainability, and privacy-preserving methods to enhance recommendation systems.

Description from the archive archive 2025-07-28.

Benchmarks archive 2025-07-28

55 leaderboard tables shown for this task, 55 with rows (a “benchmark” on this site is a table with at least one row, as on /sota), ordered by row count. “Best model” is the first row in the archive's own order at snapshot; nothing is re-ranked here and metric direction is not recorded in the archive. PwC's Trend sparklines are not in the archive, so that column is omitted. 10 shown of 55 until expanded.

Syntology 11 newer papers were judged to report on this task's benchmarks so far (Claude Sonnet 4.5, reading their tables: a model's judgement, not a result; 0 placed): MovieLens 1M (4 judged, 0 placed), MovieLens 20M (1 judged, 0 placed), Amazon-Book (2 judged, 0 placed), Gowalla (4 judged, 0 placed), Yelp2018 (3 judged, 0 placed), Netflix (1 judged, 0 placed), Amazon Beauty (1 judged, 0 placed), Amazon Games (1 judged, 0 placed), Yelp (1 judged, 0 placed), Amazon Books (1 judged, 0 placed), Ciao (1 judged, 0 placed). Drawn from 8,886 of 18,637 newer papers checked so far (733 more had nothing to place; 41 could not be read).

DatasetBest model (first row in archive order)PaperCodeSyntologyCompare
MovieLens 1M (31 rows) GLocal-K GLocal-K: Global and Local Kernels for Recommender Systems code — Compare
MovieLens 100K (18 rows) CDLD-GR (doi: 10.5281/zenodo.15851754) — — — Compare
MovieLens 20M (18 rows) HyperML HyperML: A Boosting Metric Learning Approach in Hyperbolic Space... — — Compare
MovieLens 10M (17 rows) Bayesian timeSVD++ flipped On the Difficulty of Evaluating Baselines: A Study on Recommender Systems code Syntology 3 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 0 violated, 3 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified Compare
Amazon-Book (16 rows) SSCF Sapling Similarity: a performing and interpretable memory-based... code — Compare
Gowalla (13 rows) ConvNCF Outer Product-based Neural Collaborative Filtering code — Compare
Yelp2018 (11 rows) NESCL Neighborhood-Enhanced Supervised Contrastive Learning for... code Syntology 2 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified Compare
Netflix (10 rows) H+Vamp Gated Enhancing VAEs for Collaborative Filtering: Flexible Priors &... code — Compare
Douban Monti (8 rows) GLocal-K GLocal-K: Global and Local Kernels for Recommender Systems code — Compare
ReDial (8 rows) KERL Knowledge Graphs and Pre-trained Language Models enhanced... code — Compare
Douban (7 rows) I-CFN Hybrid Recommender System based on Autoencoders code — Compare
Million Song Dataset (7 rows) EASE Embarrassingly Shallow Autoencoders for Sparse Data code Syntology 5 ran (of which 0 constructed an object rather than computing a result; 5 with no instrument failure: 0 honoured, 0 violated, 5 with no contract checked; 0 where Syntology's instrument failed) · 1 unverified Compare
Flixster Monti (7 rows) IGMC Inductive Matrix Completion Based on Graph Neural Networks code Syntology official (archive's flag): 1 ran · 1 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 1 honoured, 0 violated, 0 with no contract checked; 0 where Syntology's instrument failed) · 1 unverified Compare
Amazon Beauty (6 rows) CARCA Abs + Con Positional encoding is not the same as context: A study on... code Syntology official (archive's flag): 8 ran · 8 ran (of which 0 constructed an object rather than computing a result; 4 with no instrument failure: 3 honoured, 1 violated, 0 with no contract checked; 4 where Syntology's instrument failed) · 0 unverified Compare
Amazon Games (6 rows) ProxyRCA Proxy-based Item Representation for Attribute and Context-aware... code — Compare
YahooMusic Monti (6 rows) MG-GAT Interpretable Recommender System With Heterogeneous Information: A... code — Compare
Flixster (4 rows) GRALS Collaborative Filtering with Graph Information: Consistency and... code — Compare
Amazon Fashion (4 rows) ProxyRCA Proxy-based Item Representation for Attribute and Context-aware... code — Compare
Epinions (4 rows) DANSER Dual Graph Attention Networks for Deep Latent Representation of... code — Compare
YahooMusic (3 rows) GRALS Scalable Probabilistic Matrix Factorization with Graph-Based Priors code — Compare
Amazon Men (3 rows) CARCA Learnt + Con Positional encoding is not the same as context: A study on... code Syntology official (archive's flag): 8 ran · 8 ran (of which 0 constructed an object rather than computing a result; 4 with no instrument failure: 3 honoured, 1 violated, 0 with no contract checked; 4 where Syntology's instrument failed) · 0 unverified Compare
Last.FM (3 rows) Ekar* Ekar: An Explainable Method for Knowledge Aware Recommendation code — Compare
Polyvore (3 rows) Fashion GAE Context-Aware Visual Compatibility Prediction code — Compare
Amazon-CDs (2 rows) HetroFair Heterophily-Aware Fair Recommendation using Graph Convolutional Networks code — Compare
Amazon Product Data (2 rows) TLSAN TLSAN: Time-aware Long- and Short-term Attention Network for... code — Compare
Book-Crossing (2 rows) TransCF Collaborative Translational Metric Learning code — Compare
DBbook2014 (2 rows) KTUP (soft) Unifying Knowledge Graph Learning and Recommendation: Towards a... code — Compare
Frappe (2 rows) INN Interaction-aware Factorization Machines for Recommender Systems code — Compare
WeChat (2 rows) DANSER Dual Graph Attention Networks for Deep Latent Representation of... code — Compare
Yelp (2 rows) DGRec Session-based Social Recommendation via Dynamic Graph Attention Networks code Syntology official (archive's flag): 1 ran · 1 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 1 honoured, 0 violated, 0 with no contract checked; 0 where Syntology's instrument failed) · 12 unverified Compare
Alibaba-iFashion (1 row) HAKG HAKG: Hierarchy-Aware Knowledge Gated Network for Recommendation code — Compare
Amazon-Beauty (1 row) HetroFair Heterophily-Aware Fair Recommendation using Graph Convolutional Networks code — Compare
Amazon-book (1 row) LT-OCF LT-OCF: Learnable-Time ODE-based Collaborative Filtering code Syntology official (archive's flag): 3 ran · 3 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 0 violated, 3 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified Compare
Amazon Books (1 row) Multi-Gradient Descent Multi-Gradient Descent for Multi-Objective Recommender Systems code — Compare
Amazon C&A (1 row) TransCF Collaborative Translational Metric Learning code — Compare
Amazon-Electronics (1 row) HetroFair Heterophily-Aware Fair Recommendation using Graph Convolutional Networks code — Compare
Amazon-Health (1 row) HetroFair Heterophily-Aware Fair Recommendation using Graph Convolutional Networks code — Compare
Amazon-Movies (1 row) HetroFair Heterophily-Aware Fair Recommendation using Graph Convolutional Networks code — Compare
BeerAdvocate (1 row) CFM Predicting ratings in multi-criteria recommender systems via a... code — Compare
Ciao (1 row) TransCF Collaborative Translational Metric Learning code — Compare
CiteULike (1 row) RATE-CSE Collaborative Similarity Embedding for Recommender Systems code — Compare
Declicious (1 row) TransCF Collaborative Translational Metric Learning code — Compare
Delicious (1 row) DGRec Session-based Social Recommendation via Dynamic Graph Attention Networks code Syntology official (archive's flag): 1 ran · 1 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 1 honoured, 0 violated, 0 with no contract checked; 0 where Syntology's instrument failed) · 12 unverified Compare
Dianping-Food (1 row) KGNN-LS Knowledge-aware Graph Neural Networks with Label Smoothness... code Syntology 2 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 0 where Syntology's instrument failed) · 2 unverified Compare
Echonest (1 row) RANK-CSE Collaborative Similarity Embedding for Recommender Systems code — Compare
Epinions-Extend (1 row) RANK-CSE Collaborative Similarity Embedding for Recommender Systems code — Compare
Fashion-Similar (1 row) SR-PredAO(SGNN-HN) SR-PredictAO: Session-based Recommendation with High-Capability... code — Compare
GoodReads-Children (1 row) HGN Hierarchical Gating Networks for Sequential Recommendation code Syntology official (archive's flag): 3 ran · 3 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified Compare
GoodReads-Comics (1 row) HGN Hierarchical Gating Networks for Sequential Recommendation code Syntology official (archive's flag): 3 ran · 3 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified Compare
Last.FM-360k (1 row) RANK-CSE Collaborative Similarity Embedding for Recommender Systems code — Compare
MovieLens-Latest (1 row) RATE-CSE Collaborative Similarity Embedding for Recommender Systems code — Compare
Pinterest (1 row) TransCF Collaborative Translational Metric Learning code — Compare
PixelRec (1 row) SASRec An Image Dataset for Benchmarking Recommender Systems with Raw Pixels code Syntology official (archive's flag): 4 ran · 4 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 1 violated, 0 with no contract checked; 3 where Syntology's instrument failed) · 1 unverified Compare
Steam (1 row) SASRec Self-Attentive Sequential Recommendation code Syntology 12 ran (of which 0 constructed an object rather than computing a result; 10 with no instrument failure: 1 honoured, 0 violated, 9 with no contract checked; 2 where Syntology's instrument failed) · 8 unverified Compare
Tradesy (1 row) TransCF Collaborative Translational Metric Learning code — Compare

Syntology column: samples harvested from the paper's repositories and executed on synthesized fixtures; “ran” is not a correctness claim and does not order the table. Each line reads “N ran (of which C constructed an object rather than computing a result; K with no instrument failure: H honoured, V violated, P with no contract checked; I where Syntology's instrument failed) · U unverified”; the instrument figure counts failures of Syntology's instrument, not of the code. When the archive marks a repository official for the row's paper, the line starts with that repository's state (“official” is the archive's flag, not a verdict on who wrote the code; “community repositories only” when every sample that ran came from a community repository, “official: no sample here; runs from other or unrecorded repositories” when some came from a repository the paper names or has in its text, or from none recorded); hover it for the repositories the samples that ran came from. A dash means no Syntology record for that paper, not a recorded non-run. Each “ran” line links to that paper's sample list. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. For agents: get_harvested_code_for_paper(arxiv_id) lists each paper's samples; how to connect.

Libraries

Not in the archive: the export carries no per-task library table, so there is nothing to show at snapshot 2025-07-28.

Datasets archive 2025-07-28

56 datasets whose archive record lists this task, ordered by the archive's paper count. 30 shown of 56 until expanded.

Subtasks archive 2025-07-28

11 subtasks in the archive's task tree.

Papers archive 2025-07-28

archive papers tagged: 6,047 · with a code link: 1,997 · where Syntology ran a sample: 330 (270 with a run with no instrument failure, 60 where every run was a failure of Syntology's instrument) Syntology

Show: all tagged papersonly where code ran (330 of 6,047 tagged: 270 with a run with no instrument failure, 60 where every run was a failure of Syntology's instrument)

Page 1 of 61: papers 1 to 100 of 6,047, in archive order: by repositories listed in the archive (most first), then newest first, not by stars (the archive holds no stars, so PwC's “Social” and “Latest” sorts cannot be reproduced). Papers that list no repository come after every paper that lists one.

Papers without a page here are shown as plain text. A Syntology line reads “N ran (of which C constructed an object rather than computing a result; K with no instrument failure: H honoured, V violated, P with no contract checked; I where Syntology's instrument failed) · U unverified”; the figure “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. When the archive marks a repository official for the paper, the line starts with that repository's state (the archive's flag, not a verdict on who wrote the code); hover it for the repositories the samples that ran came from. The first sentence of the abstract is shown for the first 30 papers; every abstract is on its paper's page.

Syntology lines on 52 of the papers shown; no Syntology record for the others (a paper without an arXiv id cannot be joined to the graph, and absence from the graph layer is not a recorded non-run). “Ran” means the sample executed on a synthesized fixture, not that the paper's result was reproduced; each line links to that paper's sample list. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. For agents: get_harvested_code_for_paper(arxiv_id) lists each paper's samples; how to connect.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections