Performance Prediction for Conversational Search Using Perplexities of Query Rewrites

We consider query performance prediction (QPP) task for conversational search (CS), i.e