https://github.com/chaitjo/knowledge-graphs
Building Knowledge Graphs from Unstructured Text
https://github.com/chaitjo/knowledge-graphs
knowledge-graph networkx neuralcoref spacy unstructured-data wikipedia
Last synced: about 2 months ago
JSON representation
Building Knowledge Graphs from Unstructured Text
- Host: GitHub
- URL: https://github.com/chaitjo/knowledge-graphs
- Owner: chaitjo
- Created: 2020-04-03T16:14:10.000Z (about 5 years ago)
- Default Branch: master
- Last Pushed: 2020-04-15T13:59:16.000Z (about 5 years ago)
- Last Synced: 2023-03-05T10:08:17.591Z (about 2 years ago)
- Topics: knowledge-graph, networkx, neuralcoref, spacy, unstructured-data, wikipedia
- Language: Jupyter Notebook
- Homepage:
- Size: 42.9 MB
- Stars: 11
- Watchers: 1
- Forks: 6
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Building Knowledge Graphs from Unstructured Text
Challenge: Build a Knowledge Graph for the company **Bayer**, focused on their **Pharmacology business**.
Refer to the notebook [`main.ipynb`](main.ipynb) for usage and visualizations of our results. Check out the [presentation slides here](Chaitanya%20Joshi%20-%20Knowledge%20Graphs%20from%20Unstructured%20Text.pdf).
## Installation
```sh
# Create conda environment
conda create -n nlp python=3.7
conda activate nlp
# Install and setup Spacy
conda install -c conda-forge spacy==2.1.6
python -m spacy download en
python -m spacy download en_core_web_lg
# Install neuralcoref (specific version, for spacy compatibility)
conda install cython
curl https://github.com/huggingface/neuralcoref/archive/4.0.0.zip -o neuralcoref-4.0.0.zip -J -L -k
cd neuralcoref
python setup.py build_ext --inplace
python setup.py install
# Install additional packages
pip install wikipedia-api
conda install pandas networkx matplotlib seaborn
conda install pytorch=1.2.0 cudatoolkit=10.0 -c pytorch
pip install transformers
# Install extras
conda install ipywidgets nodejs -c conda-forge
jupyter labextension install @jupyter-widgets/jupyterlab-manager
```