GraphScope

🔨 🍇 💻 🚀 GraphScope: A One-Stop Large-Scale Graph Computing System from Alibaba | 一站式图计算系统

What is GraphScope?

🔨 🍇 💻 🚀 GraphScope: A One-Stop Large-Scale Graph Computing System from Alibaba | 一站式图计算系统

Overview of GraphScope

What it is

GraphScope is a unified distributed graph computing platform from Alibaba that provides a one-stop environment for performing diverse graph operations on a cluster of computers through a Python interface. It combines several Alibaba technologies: GRAPE, MaxGraph, and Graph-Learn (GL) for analytics, interactive, and graph neural networks (GNN) computation, respectively, and the Vineyard store for efficient in-memory data transfers. The project is open source under the Apache-2.0 license.

Key features

  • Provides a unified platform for graph analytics, interactive queries, and graph neural network (GNN) computation.
  • Supports running in standalone mode or on clusters managed by Kubernetes within containers.
  • Includes built-in algorithms for community detection, paths and connectivity, centrality, and more.
  • Allows users to write custom algorithms using the Pregel model or PIE model.
  • Offers a high-level language called Gremlin for graph traversal and interactive queries.
  • Models graph data as property graphs, where edges and vertices are labeled and have properties.
  • Provides a set of functions to load graph datasets from ogb and snap for convenience.

Use cases

GraphScope is suitable for large-scale graph data processing tasks such as node classification on citation networks, interactive graph exploration and querying, graph analytics (e.g., k-core decomposition, triangle counting), and training graph neural networks (GNNs) on heterogeneous graphs. It can be used in scenarios like scientific research, social network analysis, and recommendation systems.

Getting started

GraphScope can be installed as a Python package via pip. The following command installs the pre-compiled package for standalone mode:bashpip3 install graphscopeNote that graphscope requires Python >= 3.8 and pip >= 19.3. The package is built for Linux (Ubuntu 20.04+ / CentOS 7+) and macOS 12+ (Intel/Apple silicon). For Windows users, installing Ubuntu on WSL2 is recommended. For cluster mode on Kubernetes, see the repository README.

License and pricing

GraphScope is open source under the Apache-2.0 license. Pricing is not stated; the project is listed as open source.

Based on the project's public README and repository metadata.

Add this badge to your README

Featured on YunoDB Badge styles & help

Maintain this project? Show that it is featured on YunoDB.

Markdown[![Featured on YunoDB](https://yunodb.com/badge/alibabagraphscope.svg)](https://yunodb.com/tools/alibabagraphscope?utm_source=github&utm_medium=badge)
HTML<a href="https://yunodb.com/tools/alibabagraphscope?utm_source=github&utm_medium=badge"><img src="https://yunodb.com/badge/alibabagraphscope.svg" alt="Featured on YunoDB" height="20"></a>