Hi,
Currently our HR application build on Pega hosts Pega DMS. Presently, a user can enter any keyword and system displays list of questions for that keyword.
Application host a data type which has a column which stores questions and another column which stores list of associates keywords for that questions. System fetches corresponding questions based on keywords (different questions can have same keywords)
We would like to explore an option of Pega NLP here where instead of storing the mapping of questions and keywords in data type, we should be able to train the model using list of questions and based on keywords, Pega should return corresponding questions. In essence, Pega should be able to identify keywords from that questions.
Could anyone suggest any approach to achieve above business requirement?
@Sandip Pattalwar Pega NLP can be used to identify keywords from a list of questions and return corresponding questions based on those keywords. This can be achieved by using Pega’s text analytics capabilities. You can create a text prediction model in Prediction Studio, where you can add training data manually or as a batch. This training data should include examples of the keywords and corresponding questions. The model will then be able to identify these keywords in incoming text and return the corresponding questions. You can also use entity extraction models to detect specific entities such as keywords. Remember that machine learning models require training data to function effectively
This is a GenAI-powered tool. All generated answers require validation against the provided references.
Analyzing natural language
Pega NLP overview > Transcript
Learning natural language processing with NLP Sample > Exploring NLP features