Given a message: "Hey I am XYZ person (description about oneself), and I was thinking to launch a youtube video, wanted to get in touch with someone with similar experience", the model should detect two things: 1) description about the sender, 2) intention- "to connect with someone who has some experience in launching yt videos".
Here, the model should be able to detect what "same" implies.
I realized I can't make do with only intent classification because classification can only be done based on trained labels and if prompt changes to something like,
"Hey I want to connect with someone who lives in Europe and has a newly set up business in ai", the model has to detect multiple elements - (lives in Europe, newly set up business in ai).
How do I set up the training data and how do I go about solving this using Bert? My best solution was information extraction or a combined approach of intent classification and NER, where intent can be classified based on fixed labels- to seek collaboration/ seek advice. Etc.
