Processing graphs that evolve over time has seen renewed attention. Processing solutions on dynamic graphs (often dubbed "graph streaming" solutions) aim to maintain the state for a graph query as the graph evolves over time, and to timely offer a solution (approximate, or precise) when requested by the user. In this space, and in the context of shared-nothing platforms, solutions have been proposed only for relatively simple problems (e.g., BFS, SSSP, PageRank), and some are limited to incremental-only evolutions traces. Support for more complex problems remains rather unexplored. To close this gap, we present a solution for the maximum flow problem that supports both add and delete events. We build this solution on top of an event-based abstraction. Integral to this abstraction is that events tied to both graph topology changes and algorithmic maintenance are processed asynchronously, concurrently, and autonomously (i.e., without shared state). We show that our implementation provides favourable time-to-solution and scales well by evaluating it on a real-world dynamic graph with 80 million edges. We compare its performance with snapshot-based solutions both internally (with our own implementation of a shared-nothing static algorithm) and externally (with Galois, a popular shared-memory framework for static graphs).
{"title":"Going with the Flow: Real-Time Max-Flow on Asynchronous Dynamic Graphs","authors":"Juntong Luo, Scott Sallinen, M. Ripeanu","doi":"10.1145/3594778.3594882","DOIUrl":"https://doi.org/10.1145/3594778.3594882","url":null,"abstract":"Processing graphs that evolve over time has seen renewed attention. Processing solutions on dynamic graphs (often dubbed \"graph streaming\" solutions) aim to maintain the state for a graph query as the graph evolves over time, and to timely offer a solution (approximate, or precise) when requested by the user. In this space, and in the context of shared-nothing platforms, solutions have been proposed only for relatively simple problems (e.g., BFS, SSSP, PageRank), and some are limited to incremental-only evolutions traces. Support for more complex problems remains rather unexplored. To close this gap, we present a solution for the maximum flow problem that supports both add and delete events. We build this solution on top of an event-based abstraction. Integral to this abstraction is that events tied to both graph topology changes and algorithmic maintenance are processed asynchronously, concurrently, and autonomously (i.e., without shared state). We show that our implementation provides favourable time-to-solution and scales well by evaluating it on a real-world dynamic graph with 80 million edges. We compare its performance with snapshot-based solutions both internally (with our own implementation of a shared-nothing static algorithm) and externally (with Galois, a popular shared-memory framework for static graphs).","PeriodicalId":371215,"journal":{"name":"Proceedings of the 6th Joint Workshop on Graph Data Management Experiences & Systems (GRADES) and Network Data Analytics (NDA)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115563400","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Business Process Management algorithms are heavily limited by suboptimal algorithmic implementations that cannot leverage state-of-the-art algorithms in the field of relational and graph databases. The recent interest in this discipline for various IT sectors (cyber-security, Industry 4.0, and e-Health) calls for defining new algorithms improving the performance of existing ones. This paper focuses on generating several traces collected in a log from declarative temporal models by pre-emptively representing those as a specific type of finite state automaton: we show that this task boils down to a single-source multi-target graph traversal on such automaton where both the number of distinct paths to be visited as well as their length are bounded. This paper presents a novel algorithm running in polynomial time over the size of the declarative model represented as a graph and the desired log's size. The final experiments show that the resulting algorithm outperforms the state-of-the-art data-aware and dataless sequence generations in business process management.
{"title":"Fast Synthetic Data-Aware Log Generation for Temporal Declarative Models","authors":"Giacomo Bergami","doi":"10.1145/3594778.3594881","DOIUrl":"https://doi.org/10.1145/3594778.3594881","url":null,"abstract":"Business Process Management algorithms are heavily limited by suboptimal algorithmic implementations that cannot leverage state-of-the-art algorithms in the field of relational and graph databases. The recent interest in this discipline for various IT sectors (cyber-security, Industry 4.0, and e-Health) calls for defining new algorithms improving the performance of existing ones. This paper focuses on generating several traces collected in a log from declarative temporal models by pre-emptively representing those as a specific type of finite state automaton: we show that this task boils down to a single-source multi-target graph traversal on such automaton where both the number of distinct paths to be visited as well as their length are bounded. This paper presents a novel algorithm running in polynomial time over the size of the declarative model represented as a graph and the desired log's size. The final experiments show that the resulting algorithm outperforms the state-of-the-art data-aware and dataless sequence generations in business process management.","PeriodicalId":371215,"journal":{"name":"Proceedings of the 6th Joint Workshop on Graph Data Management Experiences & Systems (GRADES) and Network Data Analytics (NDA)","volume":"62 11","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114060256","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Graph neural networks (GNNs) are widely used in many downstream applications, such as graphs and nodes classification, entity resolution, link prediction, and question answering. Several interpretability methods for GNNs have been proposed recently. However, since they have not been thoroughly compared with each other, their trade-offs and efficiency in the context of underlying GNNs and downstream applications are unclear. To support more research in this domain, we develop an end-to-end interactive tool, named gInterpreter, by re-implementing 15 recent GNN interpretability methods in a common environment on top of a number of state-of-the-art GNNs employed for different downstream tasks. This paper demonstrates gInterpreter with an interactive performance profiling of 15 recent GNN inter-pretability methods, aiming to explain the complex deep learning pipelines over graph-structured data.
{"title":"A Demonstration of Interpretability Methods for Graph Neural Networks","authors":"Ehsan Bonabi Mobaraki, Arijit Khan","doi":"10.1145/3594778.3594880","DOIUrl":"https://doi.org/10.1145/3594778.3594880","url":null,"abstract":"Graph neural networks (GNNs) are widely used in many downstream applications, such as graphs and nodes classification, entity resolution, link prediction, and question answering. Several interpretability methods for GNNs have been proposed recently. However, since they have not been thoroughly compared with each other, their trade-offs and efficiency in the context of underlying GNNs and downstream applications are unclear. To support more research in this domain, we develop an end-to-end interactive tool, named gInterpreter, by re-implementing 15 recent GNN interpretability methods in a common environment on top of a number of state-of-the-art GNNs employed for different downstream tasks. This paper demonstrates gInterpreter with an interactive performance profiling of 15 recent GNN inter-pretability methods, aiming to explain the complex deep learning pipelines over graph-structured data.","PeriodicalId":371215,"journal":{"name":"Proceedings of the 6th Joint Workshop on Graph Data Management Experiences & Systems (GRADES) and Network Data Analytics (NDA)","volume":"901 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130757907","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Connectivity is the cornerstone of our contemporary world, permeating various sectors like retail, communications, biology, and finance. Although this inherent interconnectedness holds substantial meaning and predictive power, harnessing it for practical use in the business realm often proves challenging. In this presentation, we will delve into the commercial applications of graph analytics, highlighting both common pitfalls to avoid and promising opportunities to explore. To begin, we will explore the prevalent use cases of graph analytics, encompassing areas such as fraud detection, supply chain optimization, data management, and recommendations. We'll also shed light on why many teams tend to deploy only a limited set of graph algorithms. Additionally, we will examine how the COVID-19 pandemic has impacted the utilization of graphs in business settings. Next, we will venture into the major mistakes that businesses often make when implementing graph analytics. These blunders range from technical hurdles like scalability issues and handling tricky data types to human challenges such as fostering a graph-thinking mindset and avoiding excessive perfectionism. Moreover, you will gain quick tips to help teams secure funding for graph projects. Lastly, we will delve into some of the most significant prospects within the commercial space. We will address enduring challenges, such as transforming business data into a graph format and ensuring interoperability with production processes. We will also dedicate time to exploring the rising interest in combining graphs with AI systems, particularly the recent buzz surrounding combining graphs with generative AI. While this particular trend garners attention, we will look at other promising opportunities that it may overshadow. By the end of this talk, you will have gained a comprehensive understanding of the practical applications of graph analytics in business contexts. Furthermore, you'll gain valuable knowledge about pitfalls to avoid, strategies for securing funding, and a forward-looking perspective on emerging possibilities in this dynamic field.
{"title":"The Commercial Side of Graph Analytics: Big Uses, Big Mistakes, Big Opportunities","authors":"A. Hodler","doi":"10.1145/3594778.3596883","DOIUrl":"https://doi.org/10.1145/3594778.3596883","url":null,"abstract":"Connectivity is the cornerstone of our contemporary world, permeating various sectors like retail, communications, biology, and finance. Although this inherent interconnectedness holds substantial meaning and predictive power, harnessing it for practical use in the business realm often proves challenging. In this presentation, we will delve into the commercial applications of graph analytics, highlighting both common pitfalls to avoid and promising opportunities to explore. To begin, we will explore the prevalent use cases of graph analytics, encompassing areas such as fraud detection, supply chain optimization, data management, and recommendations. We'll also shed light on why many teams tend to deploy only a limited set of graph algorithms. Additionally, we will examine how the COVID-19 pandemic has impacted the utilization of graphs in business settings. Next, we will venture into the major mistakes that businesses often make when implementing graph analytics. These blunders range from technical hurdles like scalability issues and handling tricky data types to human challenges such as fostering a graph-thinking mindset and avoiding excessive perfectionism. Moreover, you will gain quick tips to help teams secure funding for graph projects. Lastly, we will delve into some of the most significant prospects within the commercial space. We will address enduring challenges, such as transforming business data into a graph format and ensuring interoperability with production processes. We will also dedicate time to exploring the rising interest in combining graphs with AI systems, particularly the recent buzz surrounding combining graphs with generative AI. While this particular trend garners attention, we will look at other promising opportunities that it may overshadow. By the end of this talk, you will have gained a comprehensive understanding of the practical applications of graph analytics in business contexts. Furthermore, you'll gain valuable knowledge about pitfalls to avoid, strategies for securing funding, and a forward-looking perspective on emerging possibilities in this dynamic field.","PeriodicalId":371215,"journal":{"name":"Proceedings of the 6th Joint Workshop on Graph Data Management Experiences & Systems (GRADES) and Network Data Analytics (NDA)","volume":"46 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115997436","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
T. Faltín, Vasileios Trigonakis, Ayoub Berdai, Luigi Fusco, Călin Iorgulescu, Sungpack Hong, Hassan Chafi
Query planning is essential for graph query execution performance. In distributed graph processing, data partitioning and messaging significantly influence performance. However, these aspects are difficult to model analytically, which makes query planning especially challenging. This paper introduces scouting queries, a lightweight mechanism to gather runtime information about different query plans, which can then be used to choose the "best" plan. In a pipelined, depth-first-oriented graph processing engine, scouting queries typically execute for a brief amount of time with negligible overhead. Partial results can be reused to avoid redundant work. We evaluate scouting queries and show that they bring speedups of up to 8.7× for heavy queries, while adding low overhead for those queries that do not benefit.
{"title":"Better Distributed Graph Query Planning With Scouting Queries","authors":"T. Faltín, Vasileios Trigonakis, Ayoub Berdai, Luigi Fusco, Călin Iorgulescu, Sungpack Hong, Hassan Chafi","doi":"10.1145/3594778.3594884","DOIUrl":"https://doi.org/10.1145/3594778.3594884","url":null,"abstract":"Query planning is essential for graph query execution performance. In distributed graph processing, data partitioning and messaging significantly influence performance. However, these aspects are difficult to model analytically, which makes query planning especially challenging. This paper introduces scouting queries, a lightweight mechanism to gather runtime information about different query plans, which can then be used to choose the \"best\" plan. In a pipelined, depth-first-oriented graph processing engine, scouting queries typically execute for a brief amount of time with negligible overhead. Partial results can be reused to avoid redundant work. We evaluate scouting queries and show that they bring speedups of up to 8.7× for heavy queries, while adding low overhead for those queries that do not benefit.","PeriodicalId":371215,"journal":{"name":"Proceedings of the 6th Joint Workshop on Graph Data Management Experiences & Systems (GRADES) and Network Data Analytics (NDA)","volume":"17 7","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"113957008","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Most previous research considers processing queries on the current or previous states of a graph. In this paper, we propose processing future-time graph queries, i.e., predicting the output of a query on some future state of the graph. To process future-time queries, we present a generic approach that exploits a predictive model that provides oracles about the future state of the graph. We focus on future-time shortest path queries that given a temporal graph and two nodes return the shortest path between them at some future time. We present two algorithms each invoking a different type of oracle: (a) a link prediction oracle that given two nodes returns the probability of an edge between them, and (b) a connection prediction oracle that given a node u and a future time instance t returns the node υ that u will connect to at t. Finally, we present experimental results using off-the-shelf prediction models that provide such oracles.
{"title":"Future-Time Temporal Path Queries","authors":"Christos Gkartzios, E. Pitoura","doi":"10.1145/3594778.3594879","DOIUrl":"https://doi.org/10.1145/3594778.3594879","url":null,"abstract":"Most previous research considers processing queries on the current or previous states of a graph. In this paper, we propose processing future-time graph queries, i.e., predicting the output of a query on some future state of the graph. To process future-time queries, we present a generic approach that exploits a predictive model that provides oracles about the future state of the graph. We focus on future-time shortest path queries that given a temporal graph and two nodes return the shortest path between them at some future time. We present two algorithms each invoking a different type of oracle: (a) a link prediction oracle that given two nodes returns the probability of an edge between them, and (b) a connection prediction oracle that given a node u and a future time instance t returns the node υ that u will connect to at t. Finally, we present experimental results using off-the-shelf prediction models that provide such oracles.","PeriodicalId":371215,"journal":{"name":"Proceedings of the 6th Joint Workshop on Graph Data Management Experiences & Systems (GRADES) and Network Data Analytics (NDA)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114419525","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Andrew Chai, Alireza Vezvaei, Lukasz Golab, M. Kargar, D. Srivastava, Jaroslaw Szlichta, Morteza Zihayat
We present EAGER: a tool for answering questions expressed in natural language. Core to EAGER is a modular pipeline for generating a knowledge graph from raw text without human intervention. Notably, EAGER uses the knowledge graph to answer questions and to explain the reasoning behind the derivation of answers. Our demonstration will showcase both the automated knowledge graph generation pipeline and the explainable question answering functionality. Lastly, we outline open problems and directions for future work.
{"title":"EAGER: Explainable Question Answering Using Knowledge Graphs","authors":"Andrew Chai, Alireza Vezvaei, Lukasz Golab, M. Kargar, D. Srivastava, Jaroslaw Szlichta, Morteza Zihayat","doi":"10.1145/3594778.3594877","DOIUrl":"https://doi.org/10.1145/3594778.3594877","url":null,"abstract":"We present EAGER: a tool for answering questions expressed in natural language. Core to EAGER is a modular pipeline for generating a knowledge graph from raw text without human intervention. Notably, EAGER uses the knowledge graph to answer questions and to explain the reasoning behind the derivation of answers. Our demonstration will showcase both the automated knowledge graph generation pipeline and the explainable question answering functionality. Lastly, we outline open problems and directions for future work.","PeriodicalId":371215,"journal":{"name":"Proceedings of the 6th Joint Workshop on Graph Data Management Experiences & Systems (GRADES) and Network Data Analytics (NDA)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122536547","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Graph features are crucial to many applications such as recommender systems and risk management systems. The process to obtain useful graph features involves ingesting data from various upstream data sources, defining the desired graph features for the required applications, constructing a feature engineering workflow to compute the features, and storing and managing the resulting features for downstream tasks (e.g., graph AI and graph BI) and for future reuse. To the majority of users, especially SMEs and non-tech companies, this process poses daunting challenges as it requires users to not only learn various methods (e.g., graph analytical algorithms, non-GNN graph embeddings, GNNs) to define graph features and program their computation, but also learn many infrastructures (e.g., upstream databases, downstream ML systems, graph analytics systems) to compute, manage and use the graph features in production. These challenges have significantly restricted the wider applications of graph technologies such as graph AI and graph BI currently in industry. The current solution provided by major graph database vendors (e.g., Amazon Neptune, Neo4j, Tiger-Graph) is to connect various upstream and downstream systems to their own graph database, which is used to compute and manage graph features. However, such a solution ties users to a specific graph infrastructure that may not be the preferred infrastructure and may even require them to re-develop their applications on a new infrastructure. In addition, a specific graph database or infrastructure often does not have the best performance for all workloads and certainly does not support the computation of all types of graph features. As a result, the existing solution limits users' flexibility in choosing their own infrastructure and their productivity in developing their applications. In Part 1 of this talk, I will introduce various types of graph features and their applications. Then I will present some trends in using graph databases for graph feature computation and management, analyze the limitations of the existing methods, and identify the requirements of a graph feature management solution that is practical and highly usable to average users. In Part 2 of this talk, I will introduce our ongoing project that aims at providing a highly usable graph feature platform. Our solution decouples graph feature logic specification and management (i.e., how features are defined, coded and managed) from the generation and execution of the workflow for feature computation (i.e., execution plan generation and the actual execution), so that users can flexibly select different infrastructures suitable for the computation of specific types of graph features. It also manages the upstream, downstream and feature engineering and serving infrastructures, so as to free users from tedious tasks associated with deploying infrastructures and connecting them in a feature engineering dataflow. Thus, users can focus on creating
{"title":"Graph Feature Management: Impact, Challenges and Opportunities","authors":"James Cheng","doi":"10.1145/3594778.3596882","DOIUrl":"https://doi.org/10.1145/3594778.3596882","url":null,"abstract":"Graph features are crucial to many applications such as recommender systems and risk management systems. The process to obtain useful graph features involves ingesting data from various upstream data sources, defining the desired graph features for the required applications, constructing a feature engineering workflow to compute the features, and storing and managing the resulting features for downstream tasks (e.g., graph AI and graph BI) and for future reuse. To the majority of users, especially SMEs and non-tech companies, this process poses daunting challenges as it requires users to not only learn various methods (e.g., graph analytical algorithms, non-GNN graph embeddings, GNNs) to define graph features and program their computation, but also learn many infrastructures (e.g., upstream databases, downstream ML systems, graph analytics systems) to compute, manage and use the graph features in production. These challenges have significantly restricted the wider applications of graph technologies such as graph AI and graph BI currently in industry. The current solution provided by major graph database vendors (e.g., Amazon Neptune, Neo4j, Tiger-Graph) is to connect various upstream and downstream systems to their own graph database, which is used to compute and manage graph features. However, such a solution ties users to a specific graph infrastructure that may not be the preferred infrastructure and may even require them to re-develop their applications on a new infrastructure. In addition, a specific graph database or infrastructure often does not have the best performance for all workloads and certainly does not support the computation of all types of graph features. As a result, the existing solution limits users' flexibility in choosing their own infrastructure and their productivity in developing their applications. In Part 1 of this talk, I will introduce various types of graph features and their applications. Then I will present some trends in using graph databases for graph feature computation and management, analyze the limitations of the existing methods, and identify the requirements of a graph feature management solution that is practical and highly usable to average users. In Part 2 of this talk, I will introduce our ongoing project that aims at providing a highly usable graph feature platform. Our solution decouples graph feature logic specification and management (i.e., how features are defined, coded and managed) from the generation and execution of the workflow for feature computation (i.e., execution plan generation and the actual execution), so that users can flexibly select different infrastructures suitable for the computation of specific types of graph features. It also manages the upstream, downstream and feature engineering and serving infrastructures, so as to free users from tedious tasks associated with deploying infrastructures and connecting them in a feature engineering dataflow. Thus, users can focus on creating","PeriodicalId":371215,"journal":{"name":"Proceedings of the 6th Joint Workshop on Graph Data Management Experiences & Systems (GRADES) and Network Data Analytics (NDA)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130600300","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jeroen Bollen, Jasper Steegmans, Jan Van den Bussche, Stijn Vansummeren
Graph Neural Networks (GNNs) are a form of deep learning that enable a wide range of machine learning applications on graph-structured data. The learning of GNNs, however, is known to pose challenges for memory-constrained devices such as GPUs. In this paper, we study exact compression as a way to reduce the memory requirements of learning GNNs on large graphs. In particular, we adopt a formal approach to compression and propose a methodology that transforms GNN learning problems into provably equivalent compressed GNN learning problems. In a preliminary experimental evaluation, we give insights into the compression ratios that can be obtained on real-world graphs and apply our methodology to an existing GNN benchmark.
{"title":"Learning Graph Neural Networks using Exact Compression","authors":"Jeroen Bollen, Jasper Steegmans, Jan Van den Bussche, Stijn Vansummeren","doi":"10.1145/3594778.3594878","DOIUrl":"https://doi.org/10.1145/3594778.3594878","url":null,"abstract":"Graph Neural Networks (GNNs) are a form of deep learning that enable a wide range of machine learning applications on graph-structured data. The learning of GNNs, however, is known to pose challenges for memory-constrained devices such as GPUs. In this paper, we study exact compression as a way to reduce the memory requirements of learning GNNs on large graphs. In particular, we adopt a formal approach to compression and propose a methodology that transforms GNN learning problems into provably equivalent compressed GNN learning problems. In a preliminary experimental evaluation, we give insights into the compression ratios that can be obtained on real-world graphs and apply our methodology to an existing GNN benchmark.","PeriodicalId":371215,"journal":{"name":"Proceedings of the 6th Joint Workshop on Graph Data Management Experiences & Systems (GRADES) and Network Data Analytics (NDA)","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130611363","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Proceedings of the 6th Joint Workshop on Graph Data Management Experiences & Systems (GRADES) and Network Data Analytics (NDA)","authors":"","doi":"10.1145/3594778","DOIUrl":"https://doi.org/10.1145/3594778","url":null,"abstract":"","PeriodicalId":371215,"journal":{"name":"Proceedings of the 6th Joint Workshop on Graph Data Management Experiences & Systems (GRADES) and Network Data Analytics (NDA)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115373235","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：