README.md

🤖 Neural SPARQL Machines

A Machine-Translation Approach for Question Answering over Knowledge Graphs.

IMPORTANT

If you are looking for the code for papers "SPARQL as a Foreign Language" and "Neural Machine Translation for Query Construction and Composition" please checkout tag v0.1.0-akaha or branch v1.

Install

Via pip

Coming soon!

Local setup

Clone the repository.

pip install -r requirements.txt

Example of usage

The Generator module

Pre-generated data

You can extract pre-generated data and model checkpoints from here (1.1 GB) in folders having the respective names.

Manual Generation (Alternative to using pre-generated data)

The template used in the paper can be found in a file such as Annotations_F30_art.csv. data/art_30 will be the ID of the working dataset used throughout the tutorial. To generate the training data, launch the following command.

mkdir -p data/art_30
python nspm/generator.py --templates data/templates/Annotations_F30_art.csv --output data/art_30

Launch the command if you want to build dataset seprately else it will internally be called while training.

python nspm/data_gen.py --input data/art_30 --output data/art_30

The Learner module

Now go back to the initial directory and launch learner.py to train the model.

python nspm/learner.py --input data/art_30 --output data/art_30

This command will create a model checkpoints in data/art_30 and some pickle files in data/art_30/pickle_objects.

The Interpreter module

Predict the SPARQL query for a given question it will store the detailed output in output_query.

python nspm/interpreter.py --input data/art_30 --output data/art_30 --query "yuncken freeman has architected in how many cities?"

or, if you want to use NSpM with airml to install pre-trained models, follow these steps,

1. Install airML latest version from here
2. Navigate to the table.kns here and check if your model is listed in that file.
3. Then copy the name of that model and use it with the interpreter.py as follows

python interpreter.py --airml http://nspm.org/art --output data/art_30 --inputstr "yuncken freeman has architected in how many cities?"

Use cases & integrations

• Components of the Adam Medical platform partly developed by Jose A. Alvarado at Graphen (including a humanoid robot called Dr Adam), rely on NSpM technology.
• The Telegram NSpM chatbot offers an integration of NSpM with the Telegram messaging platform.
• The Google Summer of Code program has been supporting 6 students to work on NSpM-backed project "A neural question answering model for DBpedia" since 2018.
• A question answering system was implemented on top of NSpM by Muhammad Qasim.

Publications

SPARQL as a Foreign Language (2017)

• arXiv: https://arxiv.org/abs/1708.07624

@inproceedings{soru-marx-2017,
    author = "Tommaso Soru and Edgard Marx and Diego Moussallem and Gustavo Publio and Andr\'e Valdestilhas and Diego Esteves and Ciro Baron Neto",
    title = "{SPARQL} as a Foreign Language",
    year = "2017",
    journal = "13th International Conference on Semantic Systems (SEMANTiCS 2017) - Posters and Demos",
    url = "https://arxiv.org/abs/1708.07624",
}

Neural Machine Translation for Query Construction and Composition (2018)

• NAMPI Website: https://uclnlp.github.io/nampi/
• arXiv: https://arxiv.org/abs/1806.10478

@inproceedings{soru-marx-nampi2018,
    author = "Tommaso Soru and Edgard Marx and Andr\'e Valdestilhas and Diego Esteves and Diego Moussallem and Gustavo Publio",
    title = "Neural Machine Translation for Query Construction and Composition",
    year = "2018",
    journal = "ICML Workshop on Neural Abstract Machines \& Program Induction (NAMPI v2)",
    url = "https://arxiv.org/abs/1806.10478",
}

Exploring Sequence-to-Sequence Models for SPARQL Pattern Composition (2020)

• arXiv: https://arxiv.org/abs/2010.10900

@inproceedings{panchbhai-2020,
    author = "Anand Panchbhai and Tommaso Soru and Edgard Marx",
    title = "Exploring Sequence-to-Sequence Models for {SPARQL} Pattern Composition",
    year = "2020",
    journal = "First Indo-American Knowledge Graph and Semantic Web Conference",
    url = "https://arxiv.org/abs/2010.10900",
}

Liber AI on Medium (2020)

• What is a Neural SPARQL Machine?

Contact

Questions?

• Primary contacts: Tommaso Soru and Edgard Marx.
• Neural SPARQL Machines mailing list.
• Join the conversation on Gitter.

Follow us

• Follow the project on ResearchGate.
• Follow Liber AI Research on Twitter.