Veritass 0.0.5

A python package for checking text for false statments via the Semantic Web

Veritas

(Package name is Veritass due to namespace problems) Veritas is an opensource fact checking tool. It will be able to check a statment with the help of the Semantic Web, by converting text into sparql querry questions which will then be passed into a KnowledgeGraph.

TODO:

1. Getting an input (e.g. tweet, article etc.) -> Done
2. Converting the input into questions -> Work in progress
3. Converting the questions into a Sparql query (using later FactForge as endpoint) -> Done
4. Getting the answer and formatting it into a usefull answer -> Work in progress
5. Return it to the user -> Done

Installation:

It's a package so you can import it (disclaimer: still in development).
For development:pip install -e ./Veritas (by pulling it from here) and pull yourself the tntspa repository and initalise the submodules, so you have the version of fairseq we use for development

All code can be currently found in the Veritas folder
For using it: pip install Veritass

Quickstart:

    m = modelApi.ModelApi()
    ret:list = m.translate(text="What is the biggest Mountin in the US? What is the capital of China?")
    print(ret)

Know Issues:

Often a question won't return an answer or an related answer, we are working on it. (As you can see in the example)

Extra stuff:

We use the model checkpoint and a code snippet from: https://github.com/xiaoyuin/tntspa and the included fairseq submodule.

Citation:

  author    = {Xiaoyu Yin and
               Dagmar Gromann and
               Sebastian Rudolph},
  title     = {Neural Machine Translating from Natural Language to {SPARQL}},
  journal   = {CoRR},
  volume    = {abs/1906.09302},
  year      = {2019},
  url       = {http://arxiv.org/abs/1906.09302},
  archivePrefix = {arXiv},
  eprint    = {1906.09302},
  timestamp = {Thu, 27 Jun 2019 18:54:51 +0200},
  biburl    = {https://dblp.org/rec/journals/corr/abs-1906-09302.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org}
} ```
``` @inproceedings{ott2019fairseq,
  title = {fairseq: A Fast, Extensible Toolkit for Sequence Modeling},
  author = {Myle Ott and Sergey Edunov and Alexei Baevski and Angela Fan and Sam Gross and Nathan Ng and David Grangier and Michael Auli},
  booktitle = {Proceedings of NAACL-HLT 2019: Demonstrations},
  year = {2019},
} ```