Join the conversation on Slack
Open source library based on TensorFlow that predicts links between concepts in a knowledge graph.
AmpliGraph is a suite of neural machine learning models for relational Learning, a branch of machine learning that deals with supervised learning on knowledge graphs.
Use AmpliGraph if you need to:
- Discover new knowledge from an existing knowledge graph.
- Complete large knowledge graphs with missing statements.
- Generate stand-alone knowledge graph embeddings.
- Develop and evaluate a new relational model.
AmpliGraph's machine learning models generate knowledge graph embeddings, vector representations of concepts in a metric space:
It then combines embeddings with model-specific scoring functions to predict unseen and novel links:
The new version features TensorFlow 2 back-end and Keras style APIs that makes it faster, easier to use and
extend the support for multiple features. Further, the data input/output pipeline has changed, and the support for
some obsolete models was discontinued.
See the Changelog for a more thorough list of changes.
- Intuitive APIs: AmpliGraph APIs are designed to reduce the code amount required to learn models that predict links in knowledge graphs. The new version AmpliGraph 2 APIs are in Keras style, making the user experience even smoother.
- GPU-Ready: AmpliGraph 2 is based on TensorFlow 2, and it is designed to run seamlessly on CPU and GPU devices - to speed-up training.
- Extensible: Roll your own knowledge graph embeddings model by extending AmpliGraph base estimators.
AmpliGraph includes the following submodules:
- Datasets: helper functions to load datasets (knowledge graphs).
- Models: knowledge graph embedding models. AmpliGraph 2 contains TransE, DistMult, ComplEx, HolE (More to come!)
- Evaluation: metrics and evaluation protocols to assess the predictive power of the models.
- Discovery: High-level convenience APIs for knowledge discovery (discover new facts, cluster entities, predict near duplicates).
- Compat: submodule that extends the compatibility of AmpliGraph 2 APIs to those of AmpliGraph 1.x for the user already familiar with them.
- Linux, macOS, Windows
- Python ≥ 3.8
To provision a virtual environment for installing AmpliGraph, any option can work; here we will give provide the
instruction for using venv and Conda.
The first step is to create and activate the virtual environment.
python3.8 -m venv PATH/TO/NEW/VIRTUAL_ENVIRONMENT
source PATH/TO/NEW/VIRTUAL_ENVIRONMENT/bin/activate
Once this is done, we can proceed with the installation of TensorFlow 2:
pip install "tensorflow==2.9.0"
If you are installing Tensorflow on MacOS, instead of the following please use:
pip install "tensorflow-macos==2.9.0"
IMPORTANT: the installation of TensorFlow can be tricky on Mac OS with the Apple silicon chip. Though venv can
provide a smooth experience, we invite you to refer to the dedicated section
down below and consider using conda if some issues persist in alignment with the
Tensorflow Plugin page on Apple developer site.
The first step is to create and activate the virtual environment.
conda create --name ampligraph python=3.8
source activate ampligraph
Once this is done, we can proceed with the installation of TensorFlow 2, which can be done through pip or conda.
pip install "tensorflow==2.9.0"
or
conda install "tensorflow==2.9.0"
When installing TensorFlow 2 for Mac OS with Apple silicon chip we recommend to use a conda environment.
conda create --name ampligraph python=3.8
source activate ampligraph
After having created and activated the virtual environment, run the following to install Tensorflow.
conda install -c apple tensorflow-deps
pip install --user tensorflow-macos==2.9.0
pip install --user tensorflow-metal==0.6
In case of problems with the installation or for further details, refer to Tensorflow Plugin page on the official Apple developer website.
Once the installation of Tensorflow is complete, we can proceed with the installation of AmpliGraph.
To install the latest stable release from pip:
pip install ampligraph
To sanity check the installation, run the following:
>>> import ampligraph
>>> ampligraph.__version__
'2.1.0'If instead you want the most recent development version, you can clone the repository from
GitHub, install AmpliGraph from source and checkout the develop
branch. In this way, your local working copy will be on the latest commit on the develop branch.
git clone https://github.com/Accenture/AmpliGraph.git
cd AmpliGraph
git checkout develop
pip install -e .
Notice that the code snippet above installs the library in editable mode (-e).
To sanity check the installation run the following:
>>> import ampligraph
>>> ampligraph.__version__
'2.1-dev'AmpliGraph includes implementations of TransE, DistMult, ComplEx, HolE and RotatE. Versions <2.0 also includes ConvE, and ConvKB. Their predictive power is reported below and compared against the state-of-the-art results in literature. More details available here.
| FB15K-237 | WN18RR | YAGO3-10 | FB15k | WN18 | |
|---|---|---|---|---|---|
| Literature Best | 0.35* | 0.48* | 0.49* | 0.84** | 0.95* |
| TransE | 0.31 | 0.22 | 0.50 | 0.62 | 0.66 |
| DistMult | 0.30 | 0.47 | 0.48 | 0.71 | 0.82 |
| ComplEx | 0.31 | 0.51 | 0.49 | 0.73 | 0.94 |
| HolE | 0.30 | 0.47 | 0.47 | 0.73 | 0.94 |
| RotatE | 0.31 | 0.51 | 0.43 | 0.70 | 0.95 |
| ConvE (AmpliGraph v1.4) | 0.26 | 0.45 | 0.30 | 0.50 | 0.93 |
| ConvE (1-N, AmpliGraph v1.4) | 0.32 | 0.48 | 0.40 | 0.80 | 0.95 |
| ConvKB (AmpliGraph v1.4) | 0.23 | 0.39 | 0.30 | 0.65 | 0.80 |
** Kadlec, Rudolf, Ondrej Bajgar, and Jan Kleindienst. "Knowledge base completion: Baselines strike back. " arXiv preprint arXiv:1705.10744 (2017). Results above are computed assigning the worst rank to a positive in case of ties. Although this is the most conservative approach, some published literature may adopt an evaluation protocol that assigns the best rank instead.
The project documentation can be built from your local working copy with:
cd docs
make clean autogen html
See guidelines from AmpliGraph documentation.
If you like AmpliGraph and you use it in your project, why not starring the project on GitHub!
If you instead use AmpliGraph in an academic publication, cite as:
@misc{ampligraph,
author= {Luca Costabello and
Alberto Bernardi and
Adrianna Janik and
Aldan Creo and
Sumit Pai and
Chan Le Van and
Rory McGrath and
Nicholas McCarthy and
Pedro Tabacof},
title = {{AmpliGraph: a Library for Representation Learning on Knowledge Graphs}},
month = mar,
year = 2019,
doi = {10.5281/zenodo.2595043},
url = {https://doi.org/10.5281/zenodo.2595043}
}
AmpliGraph is licensed under the Apache 2.0 License.