Beyond the Click: A Framework for Inferring Cognitive Traces in Search
2026-02-27 • Information Retrieval
Information RetrievalHuman-Computer Interaction
AI summaryⓘ
The authors point out that current user simulators for search systems mainly mimic what users do without understanding their thoughts or feelings. To fix this, they created a new way to guess users' mental states from their actions, using a theory about how people search for information and expert input. Their method helps improve predictions about how search sessions will turn out and how to help users who are struggling. They also shared annotated data and tools so others can use their approach and build better, more human-like user simulators.
User SimulatorsInformation Foraging TheoryCognitive TracesSearch SystemsInteraction LogsSession Outcome ForecastingUser Struggle RecoveryAnnotations DatasetMulti-agent SystemOpen-source Tools
Authors
Saber Zerhoudi, Michael Granitzer
Abstract
User simulators are essential for evaluating search systems, but they primarily copy user actions without understanding the underlying thought process. This gap exists since large-scale interaction logs record what users do, but not what they might be thinking or feeling, such as confusion or satisfaction. To solve this problem, we present a framework to infer cognitive traces from behavior logs. Our method uses a multi-agent system grounded in Information Foraging Theory (IFT) and human expert judgment. These traces improve model performance on tasks like forecasting session outcomes and user struggle recovery. We release a collection of annotations for several public datasets, including AOL and Stack Overflow, and an open-source tool that allows researchers to apply our method to their own data. This work provides the tools and data needed to build more human-like user simulators and to assess retrieval systems on user-oriented dimensions of performance.