Bu yardım dokümanının Türkçe versiyonu için tıklayın.
LLM Knowledge Base Tags
LLM Knowledge Base Tags allow the RAG system to search only relevant content by associating specific keywords with related entities. This way, the language model focuses on the relevant areas to produce the most accurate answers. Tags enable you to refine content searches, enhance response accuracy, and allow Jetlink Genius to answer questions based on the most relevant information.
Steps to Activate LLM Knowledge Base Tags:
- Go to Settings → LLM Knowledge Base Labels and click here.
- On the new page, click the "Add New Label" button.
- After filling in the relevant fields, click the "Save" button.
Label: This field allows you to name your tag. This categorization is solely for managing and distinguishing tags; it is not visible to the user.
Jetbot: Defines which Jetbot will use your LLM knowledge base tag. This setting is essential for accounts with multiple Jetbots.
Knowledge Base Content: Specifies the knowledge bases in which the tag will apply. Your existing knowledge base content will appear as options here, and you may select multiple options. For understanding and creating new knowledge bases, find more information here.
Entities: Allows you to set which expressions the tag will work on through our "Entities" feature. After creating your entities with synonyms, your existing entities will appear as options here. You may select multiple options. For understanding and creating new entities, find more information here.
Example: The scenario in the screenshot above works as follows:
Since the Platform entity is selected, when synonyms of the Platform entity are triggered, RAG will search only the selected knowledge bases named WhatsApp Messenger and Instagram. This way, the expression sent to Jetbot will be answered by the LLM using the most accurate knowledge bases. 👇