University of Science and Technology of China (USTC)

recommendationfamily

LightGCN

Simplified graph convolution network for collaborative filtering recommendation. Paper by He et al. (2020).

University of Science and Technology of China (USTC)China

Figure from the LightGCN paper: chart of training loss on the Gowalla dataset for NGCF variants.

Levelfamily

Classrecommendation

Parametersnot in the notes

Contextnot in the notes

Released6 Feb 2020

Opennessopen source

Runs onnot in the notes

Statusactive

R&D Topics

Click-through-rate prediction and sequential recommendationproduced by Graph neural networks and graph transformersused by

Identity

Developer: University of Science and Technology of China (USTC)
Family: not in the notes
Released: 6 Feb 2020 sourcesannouncement page
Weights published: –
Record checked: 12 Oct 2026

Where this comes from

Developer and dates from the research notes. Release date is the vendor's public announcement; weights date is when open weights were published.

Architecture and Size

Architecture: GNN Graph convolution reduced to neighbourhood aggregation
Total parameters: – sources
Active parameters: –
Checkpoint parameters: –
Context length: not in the notes
Input: table
Output: table

Where this comes from

Parameter counts as the developer states them. Where the checkpoint holds a different total, both are shown (register A16).

Openness and Access

Openness: open source
Licence: MIT License sources
Access routes: not in the notes

Where this comes from

The licence file governs over registry tags and aggregator labels.

Deployment

Runs on: not in the notes
Footprint: Licence is that of the RecBole implementation (MIT).
Runtimes: not in the notes

Where this comes from

Deployment tiers and runtimes as named in the notes. Footprint is the notes' figure, not a measurement by AURION.

Indonesia

Indonesian support: not stated
Regional languages: not in the notes
Jakarta hosting: unknown
Languages: not in the notes

Where this comes from

Indonesia fields are carried on every model; unknown is an allowed value.

Benchmark Results

The notes record no benchmark result for this model.

Lineage

The notes record no lineage for this model.

Industry Use Cases

No industry use case links this model.

Links

Developer

University of Science and Technology of China (USTC)university, China

Model

LightGCNUniversity of Science and Technology of China (USTC), recommendation

R&D TopicsClick-Through-Rate Prediction and Sequential Recommendation Graph Neural Networks and Graph Transformers

Use CasesSuper-App Recommendation and PersonalisationRetail and LogisticsUse cases of models in the same R&D topic; none is recorded for this model itself.

Who Runs It

GrabIn production, 6 Oct 2026Deployments of these use cases; none names this model.

Graph

Organisations 1

University of Science and Technology of China (USTC)university, ChinaDeveloper

R&D topics 2

Click-Through-Rate Prediction and Sequential RecommendationNeural models that predict clicks from sparse user and item features or predict the next…produced byGraph Neural Networks and Graph TransformersNeural networks that learn from nodes, edges and their features for node classification,…used by

Sources

arxiv.org/abs/2002.02126 https://arxiv.org/abs/2002.02126
github.com/RUCAIBox/RecBole https://github.com/RUCAIBox/RecBole
github.com/RUCAIBox/RecBole2.0/blob/main/model_list.md https://github.com/RUCAIBox/RecBole2.0/blob/main/model_list.md
huggingface.co/datasets/pwc-archive/methods https://huggingface.co/datasets/pwc-archive/methods
arxiv.org/html/2002.02126 https://arxiv.org/html/2002.02126

To cite this page: Project AURION, AI Use-Cases Research & Innovation, "LightGCN", https://aurion.semai.id/models/lightgcn/, records of October 2026.