This is a wrapper program/library that encapsulates a couple of NLP libraries that are popular among the AI and ML communities.

Examples have been used to illustrate the usage as much as possible. Not all the APIs of the underlying libraries have been covered.

The idea is to keep the API language as high-level as possible, so its easier to use and stays human-readable.

Libraries / frameworks covered:

• SpaCy (site | docs)
• Textacy (github | docs)

• Python 3.7.x or higher
• Docker (optional)
• Diskspace: 2-3GB

• Clone the repo: git clone https://github.com/neomatrix369/awesome-ai-ml-dl
• cd /path/to/awesome-ai-ml-dl/examples/better-nlp
• Ensure the Docker daemon is running in the background
• Run runDockerImage.sh
• Wait for the docker image to download (first time, one-off)
• Wait for the container to get started with the JuPyter notebook running
• Copy the notebook's url published in the console, should be of the form http://....?token....) - you will have to do a tiny bit of amendment to make the url look like http://localhost:8888?token....
• Go to the browser and paste the corrected url: http://localhost:8888?token...., you should see the screen as described in Jupyter Notebook.

• Clone the repo: git clone https://github.com/neomatrix369/awesome-ai-ml-dl
• cd /path/to/awesome-ai-ml-dl/examples/better-nlp/build
• Run install-linux.sh or install-macos.sh depending on the OS you are running on
• Run install-dependencies.sh
• Run cd ..
• Run jupyter-lab notebooks/jupyter/better_nlp_spacy_texacy_examples.ipynb
• Wait for the JuPyter lab to open in the browser
• You should see the screen as described in Jupyter Notebook.

In theory, the above should work for Windows as well, if run via git-bash or cgywin with all the necessary requirements installed and available - although it has not been tested, please provide feedback or fixes if you find any.

Load the notebook in directly into Google Colab.

Setup an environment needed to be able to run these programs without having to worry about the dependencies they use.

Please be aware that even though we are install only a few components, the installation process takes some time (irrespective if you are running in via your local environment or inside a docker container). Give it about 20-30 minutes depending on network bandwidth and overall machine performance. Or you can use a pre-built docker image, see Docker environment for more details.

For the brave at heart, install the dependencies in your local environment.

./install-linux.sh

./install-macos.sh

Alternatively please refer to the Docker environment section.

In principle, the install-linux.sh script should work in the cygwin or git bash environments - although it has not been tested, please raise PR with fixes if any. Alternatively please refer to the Docker environment section.

See Docker environment

See Examples and results

See Jupyter Notebook or load the notebook in directly into Google Colab