MAB-DQA decomposes queries into aspect-aware subqueries, treats them as bandit arms, and uses preliminary page reasoning as rewards to reallocate retrieval budgets, achieving 5-18% gains on four DQA benchmarks.
correct and complete
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
MAB-DQA: Addressing Query Aspect Importance in Document Question Answering with Multi-Armed Bandits
MAB-DQA decomposes queries into aspect-aware subqueries, treats them as bandit arms, and uses preliminary page reasoning as rewards to reallocate retrieval budgets, achieving 5-18% gains on four DQA benchmarks.