How to train your chatbot on websites
Adding new website sources allows your chatbot to retrieve and utilize information directly from your website. This ensures that your chatbot can provide accurate and current responses based on the latest content available on your site. Whether it's product details, FAQs, or blog posts, integrating these sources enhances the chatbot's ability to assist users effectively.
Where can i find it?
From the main ChatLab administration page, select Chatbots in the main menu, click on selected chatbot, select Training and in the left menu choose Links.
How to add URLs for scanning?
You have following options to choose from:
-
train on whole website
-
train on single address
-
**train on sitemap
**
Train on whole website
In the Address field provide the root url of your website ie. www.yourdomain.com, choose “Whole website” from the menu and click “Start scanning the webiste”.
The ChatLab should start scanning your website, you will see a progress indicator:
If you choose the option ““I want to select individual urls for training after scanning the website:”. After the website is scanned you will see the whole structure of the website, with size of the training character in each of the links.
You can check and uncheck any of them, checked links will be included in the training.
After you select the links, click “Start the Training”
The form will disappear and you will see the bar with training progress:
Train on sitemap
In the Sitemap field provide the url of the sitemap and click “Start scanning the website”
After the sitemap is scanned you should click “Start the Training”.
After training is completed you will see the list of websites you have trained.
Train on single link
In the Address field provide the link, choose “Single address” from the menu and click “Start scanning the website”.
You should see the list of website sources with the progress bar on the newly added link that eventually should turn to green status “Training completed”
View sources
The new button in the "Training" section allows customers to see what was really scanned during the training process. Thanks to this tool, we can easily identify why our page had this specific number of characters and what content was exactly scanned.
It significantly improves our control over the chatbot
If you want to learn more about training, please visit this page: