US 12,079,217 B2
Intent-aware learning for automated sample selection in interactive data exploration
Subrata Mitra, Karnataka (IN); Yash Gadhia, Mumbai Maharashtra (IN); Tong Yu, San Jose, CA (US); Shaddy Garg, Punjab (IN); Nikhil Sheoran, Chandigarh (IN); Arjun Kashettiwar, Pune Maharashtra (IN); and Anjali Yadav, Rajasthan (IN)
Assigned to Adobe Inc., San Jose, CA (US)
Filed by Adobe Inc., San Jose, CA (US)
Filed on May 11, 2022, as Appl. No. 17/741,811.
Prior Publication US 2023/0367772 A1, Nov. 16, 2023
Int. Cl. G06F 16/2455 (2019.01); G06F 16/2453 (2019.01); G06F 16/2457 (2019.01); G06F 16/2458 (2019.01); G06F 18/21 (2023.01)
CPC G06F 16/2455 (2019.01) [G06F 16/24542 (2019.01); G06F 16/2457 (2019.01); G06F 16/2474 (2019.01); G06F 18/217 (2023.01)] 18 Claims
OG exemplary drawing
 
1. A method comprising:
receiving a query in a query sequence, the query directed toward a dataset;
applying a sampling agent module to select, for the query, a sample from among samples of the dataset, wherein the sampling agent module includes an ML model trained to select respective samples for queries via intent-based reinforcement learning, and wherein selecting the sample comprises:
providing, as input to the ML model, previous queries in the query sequence and respective responses to the previous queries,
determining a computation cost as a sum of respective computation costs of previous samples selected for the previous queries in the query sequence,
providing the computation cost as further input to the ML model, and
selecting the sample based on output from the ML model;
executing the query against the sample; and
outputting a response to the query.