Browse State-of-the-Art › Link Prediction

Link Prediction

974 papers with code · 80 benchmarks · 67 datasets archive 2025-07-28

GraphsNatural Language Processing

Link Prediction is a task in graph and network analysis where the goal is to predict missing or future connections between nodes in a network. Given a partially observed network, the goal of link prediction is to infer which links are most likely to be added or missing based on the observed connections and the structure of the network.

( Image credit: Inductive Representation Learning on Large Graphs )

Description from the archive archive 2025-07-28.

Benchmarks archive 2025-07-28

80 leaderboard tables shown for this task, 80 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 80 until expanded.

Syntology 14 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): FB15k-237 (4 judged, 0 placed), WN18RR (2 judged, 0 placed), Cora (5 judged, 0 placed), Pubmed (1 judged, 0 placed), MovieLens 25M (1 judged, 0 placed), DBLP (1 judged, 0 placed), YAGO15k (1 judged, 0 placed), Drug-Drug Interactions (1 judged, 0 placed), Wikipeople (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
FB15k-237 (75 rows) NBFNet Neural Bellman-Ford Networks: A General Graph Neural Network... code Syntology official (archive's flag): 2 ran · 12 ran (of which 0 constructed an object rather than computing a result; 8 with no instrument failure: 0 honoured, 0 violated, 8 with no contract checked; 4 where Syntology's instrument failed) · 3 unverified Compare
WN18RR (75 rows) MoCoKGC MoCoKGC: Momentum Contrast Entity Encoding for Knowledge Graph Completion — — Compare
WN18 (37 rows) Inverse Model Convolutional 2D Knowledge Graph Embeddings code — Compare
FB15k (23 rows) AutoKGE AutoSF: Searching Scoring Functions for Knowledge Graph Embedding code — Compare
PCQM-Contact (18 rows) ViT-PS Learning Probabilistic Symmetrization for Architecture Agnostic... code Syntology official (archive's flag): 8 ran · 8 ran (of which 0 constructed an object rather than computing a result; 8 with no instrument failure: 0 honoured, 0 violated, 8 with no contract checked; 0 where Syntology's instrument failed) · 5 unverified Compare
YAGO3-10 (18 rows) MEIM MEIM: Multi-partition Embedding Interaction Beyond Block Term... code Syntology official: no sample here; runs from other or unrecorded repositories · 6 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; 5 where Syntology's instrument failed) · 1 unverified Compare
ICEWS05-15 (16 rows) SPA Search to Pass Messages for Temporal Knowledge Graph Completion code — Compare
ICEWS14 (16 rows) SPA Search to Pass Messages for Temporal Knowledge Graph Completion code — Compare
Wikidata5M (14 rows) MoCoKGC MoCoKGC: Momentum Contrast Entity Encoding for Knowledge Graph Completion — — Compare
Citeseer (13 rows) NESS NESS: Node Embeddings from Static SubGraphs code — Compare
Cora (13 rows) NESS NESS: Node Embeddings from Static SubGraphs code — Compare
Pubmed (13 rows) (unnamed in the archive) An Effective Graph Learning based Approach for Temporal Link... code — Compare
GDELT (11 rows) SPA Search to Pass Messages for Temporal Knowledge Graph Completion code — Compare
FB15k (10 rows) ComplEx-N3 (reciprocal) Canonical Tensor Decomposition for Knowledge Base Completion code Syntology official (archive's flag): 1 ran · 2 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; 1 where Syntology's instrument failed) · 0 unverified Compare
UMLS (10 rows) LP-BERT Multi-task Pre-training Language Model for Semantic Network Completion code — Compare
Yelp (9 rows) PEAGAT Metapath- and Entity-aware Graph Neural Network for Recommendation code — Compare
OpenBioLink (8 rows) DistMult — — — Compare
CoDEx Medium (7 rows) ULTRA Towards Foundation Models for Knowledge Graph Reasoning code Syntology official: harvested, nothing ran · 0 ran · 1 unverified Compare
MovieLens 25M (7 rows) PEAGAT Metapath- and Entity-aware Graph Neural Network for Recommendation code — Compare
WordNet (7 rows) Hyperbolic Entailment Cones Hyperbolic Entailment Cones for Learning Hierarchical Embeddings code — Compare
CoDEx Small (6 rows) ComplEx-N3-RP Relation Prediction as an Auxiliary Training Objective for... code — Compare
CoDEx Large (6 rows) ComplEx-N3-RP Relation Prediction as an Auxiliary Training Objective for... code — Compare
FB122 (5 rows) Prob-CBR Probabilistic Case-based Reasoning for Open-World Knowledge Graph... code — Compare
NELL-995 (4 rows) Prob-CBR Probabilistic Case-based Reasoning for Open-World Knowledge Graph... code — Compare
TSP/HCP Benchmark set (4 rows) TGT-Agx4 Triplet Interaction Improves Graph Transformers: Accurate... code Syntology official (archive's flag): 9 ran · 10 ran (of which 1 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 8 where Syntology's instrument failed) · 3 unverified Compare
DBLP (3 rows) GLACE Gaussian Embedding of Large-scale Attributed Graphs code — Compare
JF17K (3 rows) HAHE HAHE: Hierarchical Attention for Hyper-Relational Knowledge Graphs... code Syntology official (archive's flag): 3 ran · 3 ran (of which 3 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) · 3 unverified; every one of the 3 samples that ran constructed an object rather than computing a result Compare
KG20C (3 rows) MEI (small) Multi-Partition Embedding Interaction with Block Term Format for... code — Compare
Citeseer (biased evaluation) (2 rows) GraphStar (double weight on positive examples) Graph Star Net for Generalized Multi-Task Learning 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) · 4 unverified Compare
Cora (biased evaluation) (2 rows) GraphStar (double weight on positive examples) Graph Star Net for Generalized Multi-Task Learning 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) · 4 unverified Compare
Decagon (2 rows) Decagon Modeling polypharmacy side effects with graph convolutional networks code — Compare
Douban (2 rows) HSRL (DW) Learning Topological Representation for Networks via Hierarchical Sampling code — Compare
GPS (2 rows) Hyper-SAGNN-E Hyper-SAGNN: a self-attention based graph neural network for hypergraphs code Syntology official (archive's flag): 7 ran · 7 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 0 honoured, 0 violated, 6 with no contract checked; 1 where Syntology's instrument failed) · 4 unverified Compare
LiveJournal (2 rows) PBG (1 partition) PyTorch-BigGraph: A Large-scale Graph Embedding System code — Compare
MovieLens 1M (2 rows) Hyper-SAGNN-W Hyper-SAGNN: a self-attention based graph neural network for hypergraphs code Syntology official (archive's flag): 7 ran · 7 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 0 honoured, 0 violated, 6 with no contract checked; 1 where Syntology's instrument failed) · 4 unverified Compare
PPI (2 rows) PPPNE PPPNE: Personalized proximity preserved network embedding — — Compare
Pubmed (biased evaluation) (2 rows) GraphStar (double weight on positive examples) Graph Star Net for Generalized Multi-Task Learning 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) · 4 unverified Compare
USAir (2 rows) SEAL Link Prediction Based on Graph Neural Networks code Syntology official: not harvested · 0 ran · 2 unverified Compare
Temp8 (2 rows) HAHE HAHE: Hierarchical Attention for Hyper-Relational Knowledge Graphs... code Syntology official (archive's flag): 3 ran · 3 ran (of which 3 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) · 3 unverified; every one of the 3 samples that ran constructed an object rather than computing a result Compare
Wiki (2 rows) (unnamed in the archive) An Effective Graph Learning based Approach for Temporal Link... code — Compare
YAGO15k (2 rows) TNTComplEx (x10) Tensor Decompositions for temporal knowledge base completion 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) · 0 unverified Compare
YAGO37 (2 rows) SEEK SEEK: Segmented Embedding of Knowledge Graphs code — Compare
YouTube (2 rows) GATNE-T Representation Learning for Attributed Multiplex Heterogeneous Network 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) · 3 unverified Compare
AbstRCT - Neoplasm (1 row) ResAttArg Multi-Task Attentive Residual Networks for Argument Mining code — Compare
ACM (1 row) GLACE Gaussian Embedding of Large-scale Attributed Graphs code — Compare
AKSW-bib (1 row) KG2Vec LSTM Expeditious Generation of Knowledge Graph Embeddings code — Compare
Alibaba (1 row) GATNE-I Representation Learning for Attributed Multiplex Heterogeneous Network 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) · 3 unverified Compare
Alibaba-S (1 row) GATNE-T Representation Learning for Attributed Multiplex Heterogeneous Network 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) · 3 unverified Compare
Amazon (1 row) GATNE-T Representation Learning for Attributed Multiplex Heterogeneous Network 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) · 3 unverified Compare
Aristo-v4 (1 row) ComplEx-N3-RP Relation Prediction as an Auxiliary Training Objective for... code — Compare
CDCP (1 row) ResAttArg Multi-Task Attentive Residual Networks for Argument Mining code — Compare
Cit-HepPH (1 row) Asymmetric Transitivity Preservation ATP: Directed Graph Embedding with Asymmetric Transitivity Preservation code — Compare
Citeseer (nonstandard variant) (1 row) GLACE Gaussian Embedding of Large-scale Attributed Graphs code — Compare
COLLAB (1 row) GatedGCN-PE Benchmarking Graph Neural Networks code Syntology official (archive's flag): 3 ran · 10 ran (of which 0 constructed an object rather than computing a result; 9 with no instrument failure: 0 honoured, 0 violated, 9 with no contract checked; 1 where Syntology's instrument failed) · 13 unverified Compare
Cora (nonstandard variant) (1 row) GLACE Gaussian Embedding of Large-scale Attributed Graphs code — Compare
DDB14 (1 row) ConE Modeling Heterogeneous Hierarchies with Relation-specific Hyperbolic Cones code Syntology official (archive's flag): 2 ran · 2 ran (of which 2 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) · 0 unverified; every one of the 2 samples that ran constructed an object rather than computing a result Compare
DRI Corpus (1 row) ResAttArg Multi-Task Attentive Residual Networks for Argument Mining code — Compare
Drug-Drug Interactions (1 row) HOGCN Predicting Biomedical Interactions with Higher-Order Graph... code — Compare
Drug-target interactions (1 row) HOGCN Predicting Biomedical Interactions with Higher-Order Graph... code — Compare
FB-AUTO (1 row) BoxE BoxE: A Box Embedding Model for Knowledge Base Completion code — Compare
FB15k-237-ind (1 row) kNN-KGE Reasoning Through Memorization: Nearest Neighbor Knowledge Graph Embeddings code — Compare
FB15k (filtered) (1 row) ParTransH Efficient Parallel Translating Embedding For Knowledge Graphs code — Compare
Gene-disease interactions (1 row) HOGCN Predicting Biomedical Interactions with Higher-Order Graph... code — Compare
Gnutella (1 row) Asymmetric Transitivity Preservation ATP: Directed Graph Embedding with Asymmetric Transitivity Preservation code — Compare
GO21 (1 row) ConE Modeling Heterogeneous Hierarchies with Relation-specific Hyperbolic Cones code Syntology official (archive's flag): 2 ran · 2 ran (of which 2 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) · 0 unverified; every one of the 2 samples that ran constructed an object rather than computing a result Compare
IMDb (1 row) Event2vec Representation Learning for Heterogeneous Information Networks via... code — Compare
Last.FM (1 row) MAGNN MAGNN: Metapath Aggregated Graph Neural Network for Heterogeneous... code — Compare
MIT (1 row) HSRL (DW) Learning Topological Representation for Networks via Hierarchical Sampling code — Compare
ogbl-collab (1 row) Edge2Node Edge2Node: Reducing Edge Prediction to Node Classification — — Compare
OpenBG500 (1 row) MoCoSA MoCoSA: Momentum Contrast for Knowledge Graph Completion with... — — Compare
protein-protein interactions (1 row) HOGCN Predicting Biomedical Interactions with Higher-Order Graph... code — Compare
Pubmed (nonstandard variant) (1 row) GLACE Gaussian Embedding of Large-scale Attributed Graphs code — Compare
SINS (1 row) mlp Deep Learning in Mobile and Wireless Networking: A Survey — — Compare
Twitter (1 row) GATNE-T Representation Learning for Attributed Multiplex Heterogeneous Network 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) · 3 unverified Compare
Wiki-Vote (1 row) Asymmetric Transitivity Preservation ATP: Directed Graph Embedding with Asymmetric Transitivity Preservation code — Compare
Wikidata12k (1 row) TimePlex Temporal Knowledge Base Completion: New Algorithms and Evaluation Protocols code Syntology official (archive's flag): 3 ran · 3 ran (of which 1 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 2 where Syntology's instrument failed) · 9 unverified Compare
Wikipeople (1 row) HAHE HAHE: Hierarchical Attention for Hyper-Relational Knowledge Graphs... code Syntology official (archive's flag): 3 ran · 3 ran (of which 3 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) · 3 unverified; every one of the 3 samples that ran constructed an object rather than computing a result Compare
WN18 (filtered) (1 row) ParTransH Efficient Parallel Translating Embedding For Knowledge Graphs code — Compare
Yago11k (1 row) TimePlex Temporal Knowledge Base Completion: New Algorithms and Evaluation Protocols code Syntology official (archive's flag): 3 ran · 3 ran (of which 1 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 2 where Syntology's instrument failed) · 9 unverified Compare
YAGO39K (1 row) TransC (bern) Differentiating Concepts and Instances for Knowledge Graph Embedding 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: 2 honoured, 0 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified 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

67 datasets whose archive record lists this task, ordered by the archive's paper count. 30 shown of 67 until expanded.

Subtasks archive 2025-07-28

6 subtasks in the archive's task tree.

Papers archive 2025-07-28

archive papers tagged: 1,949 · with a code link: 974 · where Syntology ran a sample: 233 (205 with a run with no instrument failure, 28 where every run was a failure of Syntology's instrument) Syntology

Show: all tagged papersonly where code ran (233 of 1,949 tagged: 205 with a run with no instrument failure, 28 where every run was a failure of Syntology's instrument)

Page 1 of 20: papers 1 to 100 of 1,949, 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 58 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