GraphMatch: Fusing Language and Graph Representations in a Dynamic Two-Sided Work Marketplace

Mikołaj Sacha, Hammad Jafri, Mattie Terzolo, Ayan Sinha, Andrew Rabinovich
Proceedings of UniReps: the Third Edition of the Workshop on Unifying Representations in Neural Models, PMLR 322:360-373, 2026.

Abstract

Recommending matches in a text-rich, dynamic two-sided marketplace presents unique challenges due to evolving content and interaction graphs. We introduce GraphMatch, a new large-scale recommendation framework that fuses pre-trained language models with graph neural networks to overcome these challenges. Unlike prior approaches centered on standalone models, GraphMatch is a comprehensive recipe built on powerful text encoders and GNNs working in tandem. It employs adversarial negative sampling alongside point-in-time subgraph training to learn representations that capture both the fine-grained semantics of evolving text and the time-sensitive structure of the graph. We evaluated extensively on interaction data from Upwork, a leading labor marketplace, at large scale, and discuss our approach towards low-latency inference suitable for real-time use. In our experiments, GraphMatch outperforms language-only and graph-only baselines on matching tasks while being efficient at runtime. These results demonstrate that unifying language and graph representations yields a highly effective solution to text-rich, dynamic two-sided recommendations, bridging the gap between powerful pretrained LMs and large-scale graphs in practice.

Cite this Paper

BibTeX
@InProceedings{pmlr-v322-sacha26a, title = {GraphMatch: Fusing Language and Graph Representations in a Dynamic Two-Sided Work Marketplace}, author = {Sacha, Miko{\l}aj and Jafri, Hammad and Terzolo, Mattie and Sinha, Ayan and Rabinovich, Andrew}, booktitle = {Proceedings of UniReps: the Third Edition of the Workshop on Unifying Representations in Neural Models}, pages = {360--373}, year = {2026}, editor = {Fumero, Marco and Domine, Clementine and L"ahner, Zorah and Cannistraci, Irene and Zhao, Bo and Williams, Alex}, volume = {322}, series = {Proceedings of Machine Learning Research}, month = {06 Dec}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v322/main/assets/sacha26a/sacha26a.pdf}, url = {https://proceedings.mlr.press/v322/sacha26a.html}, abstract = {Recommending matches in a text-rich, dynamic two-sided marketplace presents unique challenges due to evolving content and interaction graphs. We introduce GraphMatch, a new large-scale recommendation framework that fuses pre-trained language models with graph neural networks to overcome these challenges. Unlike prior approaches centered on standalone models, GraphMatch is a comprehensive recipe built on powerful text encoders and GNNs working in tandem. It employs adversarial negative sampling alongside point-in-time subgraph training to learn representations that capture both the fine-grained semantics of evolving text and the time-sensitive structure of the graph. We evaluated extensively on interaction data from Upwork, a leading labor marketplace, at large scale, and discuss our approach towards low-latency inference suitable for real-time use. In our experiments, GraphMatch outperforms language-only and graph-only baselines on matching tasks while being efficient at runtime. These results demonstrate that unifying language and graph representations yields a highly effective solution to text-rich, dynamic two-sided recommendations, bridging the gap between powerful pretrained LMs and large-scale graphs in practice.} }
Endnote
%0 Conference Paper %T GraphMatch: Fusing Language and Graph Representations in a Dynamic Two-Sided Work Marketplace %A Mikołaj Sacha %A Hammad Jafri %A Mattie Terzolo %A Ayan Sinha %A Andrew Rabinovich %B Proceedings of UniReps: the Third Edition of the Workshop on Unifying Representations in Neural Models %C Proceedings of Machine Learning Research %D 2026 %E Marco Fumero %E Clementine Domine %E Zorah L"ahner %E Irene Cannistraci %E Bo Zhao %E Alex Williams %F pmlr-v322-sacha26a %I PMLR %P 360--373 %U https://proceedings.mlr.press/v322/sacha26a.html %V 322 %X Recommending matches in a text-rich, dynamic two-sided marketplace presents unique challenges due to evolving content and interaction graphs. We introduce GraphMatch, a new large-scale recommendation framework that fuses pre-trained language models with graph neural networks to overcome these challenges. Unlike prior approaches centered on standalone models, GraphMatch is a comprehensive recipe built on powerful text encoders and GNNs working in tandem. It employs adversarial negative sampling alongside point-in-time subgraph training to learn representations that capture both the fine-grained semantics of evolving text and the time-sensitive structure of the graph. We evaluated extensively on interaction data from Upwork, a leading labor marketplace, at large scale, and discuss our approach towards low-latency inference suitable for real-time use. In our experiments, GraphMatch outperforms language-only and graph-only baselines on matching tasks while being efficient at runtime. These results demonstrate that unifying language and graph representations yields a highly effective solution to text-rich, dynamic two-sided recommendations, bridging the gap between powerful pretrained LMs and large-scale graphs in practice.
APA
Sacha, M., Jafri, H., Terzolo, M., Sinha, A. & Rabinovich, A.. (2026). GraphMatch: Fusing Language and Graph Representations in a Dynamic Two-Sided Work Marketplace. Proceedings of UniReps: the Third Edition of the Workshop on Unifying Representations in Neural Models, in Proceedings of Machine Learning Research 322:360-373 Available from https://proceedings.mlr.press/v322/sacha26a.html.

Related Material