I face a problem as below:
- There is a novel app and need to find the most attractive chapter in previous 30 chapters of each novel
- So that we can limit a user continue unless he make payment
I didn't find such paper/idea through google .
What I thought is :
Set different payment point in different chapter, show to random user and record the CTR . Then we can use it to get payment distribution to set proper payment point . But we don't have many user now, and cost would be large .
So I wonder is there any NLP way to detect most attractive chapter ?