How does Beth.ai understand my comments? Before engaging with this AI, I’d like better insight on how my interactions with it are undertaken.
Beth listens to your comments, which can be a search, a question, or a request to change facts (if you have Editor permissions). A search will return one or more cards; a question will do the same unless it relates to the current card - in which case a language model will answer based on the current concept card.
A request to change facts is a bit more complex, but in essence you make an assertion. I can point you to more on this if you need.
OK. Yes I wnat to make changes to the graph using comments.
You can change facts using the English Editing Commands by addressing your assertions to Beth. These assertions must be in the format of subject ~ verb phrase ~ object. The verb phrase is the section that has a verb in it, such as “is a” or “is connected to”. The object is the word, name, or ID of the concept you want to connect to.
For example:
- “beth this is a knowledge graph tool”
- “I assert that this depends on GraphBase”
You can also remove an existing fact by using the keyword NOT, like: “ia this does not use RDF”.
You can also edit existing concepts by using beth.ai English editing comment, adding a label, or changing the text in a name or digest.
So it’s just subject-predicate-object as an english sentence?
Yes, that’s right. First Order Logic is a common and useful pattern so that’s what Workspace uses.