I am currently implementing a closed-domain FAQ chatbot using https://www.sbert.net/index.html as my main model for answering questions from the users. I wish to extend the functionality of the chatbot by allowing it to identify context from a user's query, and make use of the context in the subsequent conversation.
To achieve the above functionality, I am hoping to achieve the following:
Identify questions that need to be augmented with the context Examples of such questions are "How to apply" and "What is the fee", which essentially lacks a subject. Augment the question with the existing context" For example, Inputs: (1) question = "How to apply" and (2) context = "Machine Learning" will give an output: augmented_question = "How to apply for Machine Learning" The resulting augmented_question should be syntactically and semantically accurate.
Does anyone know of any existing AI/Deep Learning/Models that can be used to achieve the above two tasks?
