Visual Editor for NLP rules

Visual Editor for creating NLP rules.

How to use

1. Watch a live demo of the NLP editor, and learn more about our future plans in our recent IBM Data Science Community presentation.

2. Walk through our Tutorial.

3. Try the editor by following the instructions below.

Try the editor

1. Clone the repository

   git clone [email protected]:CODAIT/nlp-editor.git
   

2. Navigate to the source code

   cd nlp-editor
   

The application users a NodeJS server file as proxy, this makes it easy to replace and embed the UI with any other server - Websphere, Nginx, etc.

Prerequisites

On a Terminal window, install the Node Version Manager (nvm) as follows:

curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.39.0/install.sh | bash

Reference: https://github.com/nvm-sh/nvm#installing-and-updating

Verify nvm installed properly

> nvm -v
0.39.2

Next, install the required NodeJS version; currently at 18.12.0

nvm install v18.12.0

Verify node and npm installed properly

> node -v
v18.12.0
> npm -v
8.19.2

Run the editor locally (without a backend runtime)

1. Install the dependencies

   npm install
   

2. Build the app

   npm run build
   

3. Run the app

   npm run serve
   

4. Open http://localhost:8080 in a web browser.

Run the editor locally using the IBM Watson Discovery Backend container

Additional Prerequsites:

• Docker
• IBM Watson Discovery Backend container 01-ibm_watson_discovery_web_nlp_tool_backend-<date>.tar.gz supplied to you

1. Follow steps above to Run the editor locally

2. Extract 01-ibm_watson_discovery_web_nlp_tool_backend-<date>.tar.gz into a folder of your choice, say watson_nlp_web_tool

3. Build the container image

   cd watson_nlp_web_tool
   docker build -t watson_nlp_web_tool:1.0 .
   

4. Run the container image with volumes mapped. Note that /path/to/nlp-editor is the absolute path to the nlp-editor repository (from Step 1).

   docker run -d -v /path/to/nlp-editor/Seer-Core/aql-processor/user-data-in:/app/Seer-Core/aql-processor/user-data-in -v /path/to/nlp-editor/Seer-Core/aql-processor/run-aql-result:/app/Seer-Core/aql-processor/run-aql-result --name watson_nlp_web_tool watson_nlp_web_tool:1.0
   

5. Open http://localhost:8080 in a web browser, or use reuse session from Step 1.

6. Create your NLP model. Use the Tutorial for guidance.

7. When you are satisfied with your model, click Export. A .zip file is generated on your local file system.

8. In Watson Discovery on CloudPak for Data, apply the model by following the steps in Advanced Rules Models.

Getting Help

We welcome your questions, ideas, and feedback. Please create an issue or a discussion thread.

Contributing to the NLP editor

If you are interested in helping make the NLP editor better, we encourage you to take a look at our Contributing page.