think Thailand
  • Journey to Data & AI Workshop
  • Prerequisite
    • IBM Cloud
  • IBM Watson Studio Auto AI
    • Watson Studio Overview
    • Step 1: Watson Service Creation
    • Step 2: Build and train the model
    • Step 3: Deploy the trained model
    • Step 4: Test the deployed model
  • IBM Watson Discovery
    • Discovery Overview
    • Step 1: Create Discovery Service
    • Step 2: Launch the tooling
    • Step 3: Create a collection
    • Step 4: Download the sample document and upload to your collection
    • Step 5: Querying the dataset
  • Links
    • Sample Application: Use the Watson Discovery Service to analyze cyber security breaches
    • Sample data set source
    • Preparing CSV data set to Watson Discovery Service
    • Watson Studio Documentation
    • IBM Cloud Documentation
    • Discovery API documentation
    • IBM Developer
    • Discovery documentation
    • Sample Codes: IBM Developer Code Patterns
    • Free courses: COGNITIVE CLASS.AI
Powered by GitBook
On this page

Was this helpful?

  1. IBM Watson Discovery

Step 3: Create a collection

Your first step in the Discovery tooling is to create a data collection.

PreviousStep 2: Launch the toolingNextStep 4: Download the sample document and upload to your collection

Last updated 5 years ago

Was this helpful?

A collection is a set of your documents. Why would I want more than one collection? There are a few reasons:

  • You might want multiple collections in order to separate results for different audiences.

  • The data might be so different that it doesn't make sense for it all to be queried at the same time.

The public, pre-enriched Discovery News data collection is also available for your use. It is ready to query, and you can begin to create queries on it immediately. You cannot adjust its configuration or add documents to Discovery News.

  1. Click on Upload your own data and choose Create.

  2. When your environment is ready, click the Upload your own data button, then you can Name your new collection. Name your collection InstallDocs.

    When creating a collection, under Advanced, you have the option to choose a configuration file named Default Contract Configuration. This configuration supports only the Element Classification enrichment, which can be used to extract party, nature, and category from elements in PDFs. See for details. Do not choose this option for this tutorial.

Note: You can also crawl Box, Salesforce, Microsoft SharePoint Online, IBM Cloud Object Storage, and Microsoft SharePoint 2016 data sources, or do a web crawl with the Discovery tooling. Click the Connect a data source button and see for more information.

Connecting to data sources
Element Classification