One of the most powerful training methods is through URLs. By selecting the “Websites” option and entering a URL, the platform will crawl the site starting from the provided URL, identifying all sub-pages linked beneath it. The system will extract content from these URLs and use the information to enrich the knowledge base. Here’s how to get started:

  1. Navigate to the ‘Data’ tab.

  2. Under ‘Sources’, click on ‘Websites’.

  1. Select ‘Add Website’ and input the URL you want to train from.

  2. Toggle the ‘Import Sub-Pages’ button to include all linked sub-pages.

  3. Click the ‘Arrow’ to proceed.

  1. A list of URLs will appear—use the checkboxes to select specific URLs or check all of them.

  2. Finally, click ‘Train All URLs’ to integrate the content into your bot’s knowledge base.

This method ensures your bot has access to a comprehensive dataset directly from your website.

​
Retraining URLs

If you’ve updated content on a webpage and want the chatbot to relearn from it, follow these steps:

  1. Select the URL(s) to Retrain

    • Find the list of URLs that have been added to your chatbot’s knowledge base.

    • Tick the checkbox next to the URL(s) you want to retrain.

  2. Click “Retrain”

    • Once you’ve selected the URLs, click the “Retrain” button.

    • The system will then fetch the updated content from the webpage and incorporate it into your chatbot’s responses.

​
Deleting URLs

If you want to remove outdated or incorrect sources, follow these steps:

  • 1. Select the URL(s) to Delete- Tick the checkbox next to the URL(s) you want to remove.

  • 2. Click “Delete”- Once selected, click the “Delete” button to remove the URLs from your chatbot’s training data.

TextAnalytics