Information regarding the knowledgebase-intents and recommendations for usage.
Knowledgebase-intents in general
Knowledgebase-intents serve as alternatives to the classical, editorial intents. This new form of intents enables the display of individual and dynamic responses. The process bases on generative AI/a large language model.
Compared to the responses displayed in editorial intents, the responses displayed by knowledebase-intents are not generically-repetitive. They react dynamically and individually to the specific enquiry submitted by the users.
While the responses displayed by editorial intents are based on the priorly edited response-content, the AI based knowledgebase-intents use specifically chosen sources. These sources build the knowledgebase for the then workin AI-model, generating the response.
At the moment it is possible to submit websites and pdf-documents as named sources. Note, that the sources permanently underlie the redactional authority. However, this does not apply to the specific output in the chat.
The knowledgebase-intents are activated via the menu in the editor. Klick on knowledgebase and select the option to change the intent.
To use the feature a new general contract of use (GCU) between service provider and service recipient has to be closed. All necessary information can be received by contacting our customer-succes-management.
When the intent is changed from editorial to knowledgebase, the whole content is lost. It is therefore necessary to save the content of the editorial intent prior to the change. Only this way the restoration via the change log is possible later on.
Resource management
After the changeover, the necessary resources for the knowledgebase-intents need to be added.
The resources are formative for the displayed responses in the live-chatbot. As soon as a source is not up to date anymore, it has to be changed, to prevent the display of outdated information.
The resources can be added directly in the editor of an intent or via the central tool for resource management.
Decentralized resource management (edit in intent)
Click on the "Add" button on the right-hand side to open the input mask. Adding is done by entering the link (if it is a URL) or by upload/drag-and-drop (for a PDF file).
The added source(s) appear in the source list. If there are several sources, it is possible to filter by source type. The filter is set by clicking on the PDF or website button. The sources in the topic list form the basis for the answers then played out in the live chatbot.
Centralized Resource management (coming soon)
To open the tool klick knowledgebase in the section editor.
Here all resources of all knowledgebase-intents are managed centrally.
To add new resources, klick add. Now add the URL or PDF-files as explained above.
Different to the resources added directly via the editor, the resources added here are not connected to any intent. They have to be added manually, to be accessible for the intents.
To assign a resource to an intent, click on the added resource. Select the option add intent and set the connection.
After adding the resource, the intent(s) to which it is connected are displayed in the column used in intents.
By clicking the button the editor of the intent is opened.
If a resource is not needed anymore, it can be deleted. It is necessary to delete the set connections, to delete the resource.
As soon as the connection is deleted, the button delete resource is colored red and the resource can be deleted.
A resource must not be deleted. It is sufficient to delete the connection of a resource and the intent, to delete it from the knowledgebase. Does the specific resource gain relevance once more, it is sufficient to reactivate the connection.
Playground
You can test the AI in the Playground. Questions entered here are processed by the AI and it plays out answers that would also be played out in the chatbot. The AI also indicates here which queries it does not know the answer to or for which no answer can be generated from the inserted sources.
Answers can still be rated positively. The positive evaluation encourages the AI so that this answer will be displayed for similar/identical enquiries in the future.
Knowledgebase topics: Best and worst use cases
Knowledgebase intents are primarily used to reduce the editorial effort involved in creating answers. They are therefore generally suitable in areas where high editorial effort meets low enquiry volumes (so-called long-tail enquiries).
For intents with a higher volume of enquiries, an evaluation of the response documentation is necessary beforehand. If the documentation is good, the knowledgebase topics are also suitable here.
Knowledgebase intents are also suitable where the answers are very repetitive and good documentation exists at the same time.
However, knowledgebase topics are unsuitable where the answers are dynamic and based on complex documentation, e.g. questions about individual product KPIs.
Knowledgebase topics are also unsuitable if the answer contains sensitive information, e.g. for dosages of dietary supplements or for information about the product.
Finally, a knowledgebase topic is not recommended if only minimal editorial effort is required to create an answer.
It is possible and advisable to point out to users at the beginning that the chatbot plays out the answers based on/with the help of generative AI.