GraphLab

GraphLab

The evolution of recommender systems: From the beginning to the Big Data era

Published in Matthias Dehmer, Frank Emmert-Streib, Frontiers in Data Science, 2017

Beatrice Paoli, Monika Laner, Beat Tödtli, Jouri Semenov

Dato, formerly GraphLab, is a standalone product that can be connected with Hadoop for graph analysis and machine-learning tasks. It was fully open source, but in late 2014, they transitioned into a commercial product. Their C++ processing engine Dato Core [64] has been released to the community on Github along with their interprocess communication library (for translating between C++ and Python) and graph analytics implementations. Their MLlibs are unavailable outside their enterprise packages. Dato GraphLab comprises a scalable machine-learning toolkit that includes implementation, for example, Deep learning, factor machines, topic modeling, clustering, and nearest neighbors. Distributed processing on Hadoop enables large-scale learning.

An Analysis of Distributed Programming Models and Frameworks for Large-scale Graph Processing

View Article

Journal Information

Published in IETE Journal of Research, 2022

Alejandro Corbellini, Daniela Godoy, Cristian Mateos, Silvia Schiaffino, Alejandro Zunino

The main objective of GraphLab is to provide an efficient support for Machine Learning and Data Mining algorithms, such as collaborative filtering or unsupervised classification algorithms. In this context, GraphLab provides a set of improvements over Pregel or MR, such as asynchronous computation, prioritised scheduling (to compute certain parameters before others) and serializability.

Explore chapters and articles related to this topic

The evolution of recommender systems: From the beginning to the Big Data era

An Analysis of Distributed Programming Models and Frameworks for Large-scale Graph Processing