Arabesque is a distributed graph mining system that enables quick and easy development of graph mining algorithms, while providing a scalable and efficient implementation that runs on top of Hadoop/Spark.
Benefits of Arabesque:
- Simple intuitive API, tuned for Graph Mining Problems
- Handles all the complexity of Graph Mining Algorithms transparently
- Highly Scalable
- Efficient implementation: negligible overhead compared to equivalent centralized solutions
- Support of large graphs with over a billion edges. It can process trillion of subgraphs in a commodity cluster.
- Runs as an Apache Spark or Giraph Job.
- We have a new Spark 2.0 based version(To be released in June 2017)download page.
- "Graph Data Mining with Arabesque", SIGMOD 2017 demo paper
Check our SOSP 2015 paper that describes the system.
Arabesque can be used in one of 2 ways:
- Standalone - If you simply want to test/execute one of the example algorithms included in Arabesque, you can simply download the JAR file from our download page and use the supporting execution scripts to run your desired algorithm.
- Framework - If you want to code your own graph mining algorithm on top of Arabesque, the easiest way to accomplish this is by forking our Arabeque skeleton project and follow the instructions in the README file.
Arabesque is open-sourced under the Apache 2.0 license.
The source code can be accessed from github.