Jamshid Mozafari
University of Innsbruck
Focuses on QA, information retrieval, and human-centered AI, including hint evaluation frameworks.
From Answers to Assistance: Human-in-the-Loop Question Answering in GenAI Era
This tutorial presents HLQA as a collaborative QA paradigm that supports reasoning through adaptive questioning, hinting, and explanation.
Instructor Team
Anubhav Jangra
Columbia University
Works on personalized NLG, educational AI, and human-centered reasoning support.
Smaranda Muresan
Columbia University
Researches responsible NLP, explainability, and computational methods for human-AI collaboration.
Ekaterina Kochmar
MBZUAI
Studies NLP for intelligent tutoring systems and adaptive educational technologies.
Adam Jatowt
University of Innsbruck
Specializes in information retrieval, text mining, and explainable QA with LLMs.
Tutorial Outline
A stepwise path from foundational context to deployable human-centered QA systems.
01
Evolution of QA
From retrieval and extractive systems to LLM-based reasoning.
Context
02
Cognitive Foundations
Cognitive load, scaffolding, metacognition, and learner engagement.
Theory
03
Adaptive Questioning
Socratic and learner-aware question generation strategies.
Method
04
Hinting
Progressive, non-revealing guidance for intermediate reasoning.
Method
05
Explanation
Explanations for understanding, reflection, and long-term learning.
Method
06
Personalization
User modeling and adaptive orchestration in applied settings.
Application
07
Interactive Systems
Demonstrations across educational, research, and cognitive-support settings.
Application
08
Open Challenges
Evaluation gaps, ethics, and roadmap directions for HLQA.
Future
Session Schedule
QA evolution and the transition from answer-centric systems to collaborative cognitive support.
Cognitive load, scaffolding, and metacognitive principles for human-centered QA design.
Socratic prompting, decomposition, and personalized question generation with evaluation criteria.
Hinting techniques, quality metrics, and empirical evidence on educational effectiveness.
Pedagogical explanation patterns that improve understanding and support critical thinking.
User modeling and knowledge-tracing driven adaptation across learner profiles and goals.
Educational and research-assistance systems implementing HLQA principles in practice.
Roadmap for evaluation, ethics, multilingual settings, and robust collaborative QA benchmarks.
Selected References
Survey
Yue, Z. (2025). Survey of LLM-based reasoning and modern QA systems.
Cognitive Science
Risko, E. F., & Gilbert, S. J. (2016). Cognitive offloading foundations and implications.
Hinting
Jatowt, A. et al. (2023). Hint generation methods for reasoning-oriented QA interaction.
Explainable QA
Fan, Y. et al. (2025); Schuff, H. et al. (2020). Explanation quality and pedagogical value.