README.md

DeepSearch NLP Annotator API Example

This repository contains a simple annotator fulfilling the interface required by CPS. It is meant to be cloned, and then adapted to specific models.

Example models

In terms of functionality, the current code implements several different models, ranging from very simple (dictionary look-ups) to more SOTA (spaCy-models).

These models are:

• SimpleTextGeographyAnnotator
• ScispacyBiomedAnnotator
• WatsonHealthAnnotator

Install

To install the dependencies, execute the following commands in the deepsearch-nlp-annotator-api-example folder

python3 -m venv venv
source ./venv/bin/activate
pip install -r requirements.txt

Run API server

To launch the Rest-API server, execute the following command

sh launch_server.sh

The application is exposing an interactive Swagger UI interface, which is reachable at http://localhost:5000/api/v1/ui/. By default the API requires test 123 as authentication (see green Authorize button on right top corner).

Query API server

A detailed description to query the Rest-API can be found here. We also have an easy interactive query script, which allows you to read strings from a txt-file and visualise the output. For that, first launch the server (see above) and then go to the tests folder and execute,

python interactive_test.py

By default, it will read line-by-line a TXT file. In case you want to run it on other TXT files, simply type

python interactive_test.py --filename <path-to-text-file>

In case you want to run on CCS document files, simply type

python interactive_test.py --documents <directory-with-CCS-converted-doc's>

Further details

• Running the Annotator Application through Curl
  • How to launch locally
  • Example input and output
• Use the Annotator in CPS
  • How to configure the model in CPS
• Using an Ngrok Tunnel
  • Example of local development connected to CPS
• Deploying the application
  • How to deploy for production usage
• Develop a new annotator
  • How to add your own annotator