[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-119852-en":3,"doc-seo-119852-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},119852,16904993612988,"Olivia Brown","https://ap-avatar.wpscdn.com/davatar_a8503ba1806abce46bf441b54a3ca4cd",8,"Research & Report","Temporal Graph Benchmark for Machine Learning on Temporal Graphs - research and evaluation of temporal graph models","Temporal Graph Benchmark (TGB) provides a collection of challenging, diverse benchmark datasets designed for realistic, reproducible, and robust evaluation of machine learning on temporal graphs. The datasets span multiple years, support both node-level and edge-level prediction tasks, and cover domains such as social, trade, transaction, and transportation networks. TGB defines evaluation protocols grounded in realistic use cases, benchmarks common models across datasets, and shows performance can vary substantially. For dynamic node property prediction, simpler approaches can outperform existing temporal graph models. TGB also supplies an automated pipeline with public datasets, loaders, example code, evaluation setup, and leaderboards, and remains regularly maintained with community feedback.","arXiv :2307 .01026v2 [ cs .LG] 27 Sep 2023  \nTemporal Graph Benchmark for Machine Learning on Temporal Graphs  \nShenyang Huang1,2, ∗ Farimah Poursafaei1,2, ∗ Jacob Danovitch1,2 Matthias Fey3 Weihua Hu3 Emanuele Rossi4 Jure Leskovec7 Michael Bronstein8  \nGuillaume Rabusseau 1,5,6 Reihaneh Rabbany 1,2,6  \n1Mila-Quebec AI Institute, 2 School of Computer Science, McGill University  \n3 Kumo.AI, 4Imperial College London, 5DIRO, Université de Montréal  \n6 CIFAR AI Chair, 7 Stanford University, 8University of Oxford  \nAbstract  \nWe present the Temporal Graph Benchmark (TGB), a collection of challenging and diverse benchmark datasets for realistic, reproducible, and robust evaluation of machine learning models on temporal graphs. TGB datasets are of large scale, spanning years in duration, incorporate both node and edge-level prediction tasks and cover a diverse set of domains including social, trade, transaction, and transportation networks. For both tasks, we design evaluation protocols based on realistic use-cases. We extensively benchmark each dataset and find that the performance of common models can vary drastically across datasets. In addition, on dynamic node property prediction tasks, we show that simple methods often achieve superior performance compared to existing temporal graph models. We believe that these findings open up opportunities for future research on temporal graphs. Finally, TGB provides an automated machine learning pipeline for reproducible and accessible temporal graph research, including data loading, experiment setup and performance evaluation. TGB will be maintained and updated on a regular basis and welcomes community feedback. TGB datasets, data loaders, example codes, evaluation setup, and leaderboards are publicly available at [https://tgb.complexdatalab.com/](https://tgb.complexdatalab.com/) .  \n1 Introduction  \nMany real-world systems such as social networks, transaction networks, and molecular structures can be effectively modeled as graphs, where nodes correspond to entities and edges are relations between entities. Recently, significant advances have been made for machine learning on static graphs, led by the use of Graph Neural Networks (GNNs) [22, 43, 7] and Graph Transformers [34, 23, 12], and accelerated by the availability of public datasets and standardized evaluations protocols, such as the widely adopted Open Graph Benchmark (OGB) [17] .  \nHowever, most available graph datasets are designed only for static graphs and lack the fine-grained timestamp information often seen in many real-world networks that evolve over time. Examples include social networks [32], transportation networks [8], transaction networks [38] and trade networks [33] . Such networks are formalized as Temporal Graphs (TGs) where the nodes, edges, and their features change dynamically.  \nA variety of machine learning approaches tailored for learning on TGs have been proposed in recent years, often demonstrating promising performance [35, 9, 40, 25, 20] . However, Poursafaei et al. [33] recently revealed an important issue: these TG methods often portray an over-optimistic performance  \n∗Equal contributions  \n37th Conference on Neural Information Processing Systems (NeurIPS 2023) Track on Datasets and Benchmarks.  \n—meaning they appear to perform better than they would in real-world applications—due to the inherent limitations of commonly used evaluation protocols.  \nThis over-optimism creates serious challenges for researchers. It becomes increasingly difficult to distinguish between the strengths and weaknesses of various methods when their test results suggest similarly high performance. Furthermore, there is a discrepancy between real-world applications of TG methods and the existing evaluation protocols used to assess them. Therefore, there is a pressing need for an open and standardized benchmark that enhance the evaluation process for temporal graph learning, while being aligned with real-world applications.  \nIn th","cbCaicxHqrAeaOIU","https://ap.wps.com/l/cbCaicxHqrAeaOIU","pdf",690262,1,20,"English","en",105,"# Introduction\n## Temporal graphs and real-world motivation\n## Limitations of existing temporal graph benchmarks\n## Temporal Graph Benchmark (TGB) overview\n## Dataset selection and diversity","[{\"question\":\"What is the Temporal Graph Benchmark (TGB)?\",\"answer\":\"TGB is a benchmark suite of large-scale, diverse datasets for realistic, reproducible, and robust evaluation of machine learning models on temporal graphs.\"},{\"question\":\"What prediction tasks and domains does TGB support?\",\"answer\":\"TGB supports both node-level and edge-level prediction tasks and covers domains including social, trade, transaction, and transportation networks.\"},{\"question\":\"Why do researchers need TGB’s evaluation protocols?\",\"answer\":\"Common temporal-graph evaluation protocols can lead to over-optimistic results, making it hard to distinguish method strengths and mismatching real-world application settings.\"}]","Temporal Graph Benchmark for Machine Learning on Temporal Graphs - research and evaluation of temporal graph models | PDF",1785726660,50,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"temporal-graph-benchmark-for-machine-learning-on-temporal-graphs-research-and-evaluation-of-temporal-graph-models","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/temporal-graph-benchmark-for-machine-learning-on-temporal-graphs-research-and-evaluation-of-temporal-graph-models/119852/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-03",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What is the Temporal Graph Benchmark (TGB)?","Question",{"text":75,"@type":76},"TGB is a benchmark suite of large-scale, diverse datasets for realistic, reproducible, and robust evaluation of machine learning models on temporal graphs.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What prediction tasks and domains does TGB support?",{"text":80,"@type":76},"TGB supports both node-level and edge-level prediction tasks and covers domains including social, trade, transaction, and transportation networks.",{"name":82,"@type":73,"acceptedAnswer":83},"Why do researchers need TGB’s evaluation protocols?",{"text":84,"@type":76},"Common temporal-graph evaluation protocols can lead to over-optimistic results, making it hard to distinguish method strengths and mismatching real-world application settings.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,114,119,122,126,129,133],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":29,"slug":113},6,"Technology","technology",{"id":115,"doc_module":4,"doc_module_name":46,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":21,"slug":125},9,"Religion & Spirituality","religion-spirituality",{"id":21,"doc_module":4,"doc_module_name":46,"category_name":127,"show_sort_weight":21,"slug":128},"World Cup","world-cup",{"id":130,"doc_module":4,"doc_module_name":46,"category_name":131,"show_sort_weight":130,"slug":132},10,"Lifestyle","lifestyle",{"id":134,"doc_module":4,"doc_module_name":46,"category_name":135,"show_sort_weight":106,"slug":136},19,"General","general"]