Web URLs and Links in Stack
Enhance your bot's capabilities by training it using web URLs. This guide will walk you through the process and provide key insights to consider.
Contents
Training Your Bot with Web URLs
- Navigate to Bot Training in your Stack account.
- Enter the full URL (including
https://) and select a web crawling mode (explained below). - Allow time for the URLs to be fetched and crawled.
- Choose the URLs you wish to train on and click Train Bot.
Each URL will be processed and added to the Uploaded Links Table with its status. Ensure all URLs are trained before utilizing the bot.
URL Crawling Modes
Exact URL
For precise training, use the Exact URL method. The bot will only crawl and learn from the specific URL you provide.
Steps:
- Select Exact URL.
- Enter your desired URL and click Get Data.
- The bot will crawl the URL, train itself, and update the Uploaded Links Table.
All URLs in this Domain
Expand your bot's knowledge by training it with all pages within a specific domain. This mode allows you to select which URLs to include in the training.
Steps:
- Select All URLs in this Domain.
- Enter the domain URL and click Get Data.
- Once the pages load, choose the relevant URLs for training and click Train Bot.
During selection, you'll see:
- New Pages: URLs not yet part of the bot's training. Adding them will update the Uploaded Links Table.
- Existing Pages: URLs already in the bot's dataset. Selecting these will refresh their data.
All URLs with this Path
This mode allows the bot to crawl all pages that match a specific path within the URL. The steps are similar to "All URLs in this Domain".
Uploaded Links Table
The Uploaded Links Table displays all URLs the bot has been trained on. You can refresh or delete URLs to manage the bot's knowledge base effectively.
URL Statuses:
- Getting Data: The bot is currently refreshing the URL's data.
- Trained: The bot has successfully learned from this URL. Check the "Last data refreshed at" timestamp for updates.
- Failed: Training was unsuccessful. You can retry by refreshing or remove the URL.
Note: URLs take time to appear in the Uploaded Links Table. Wait for all URLs to be listed before using the bot.
Regularly reviewing and updating your URLs will enhance the bot's performance by ensuring it responds with the most relevant information.