Frank E. Pollick

University of Glasgow
1 previous affiliation
5Machine learning papers tracked by Lacuna
Is this you?

Research Overview

Frank E. Pollick focuses on the intersection of neuroimaging, cognitive science, and information retrieval, investigating how complex cognitive states physically manifest in the brain. His trajectory spans electroencephalography studies of relevance judgment to functional magnetic resonance imaging research on the emergence of information needs. Most recently, his work explores the proactive prediction of these needs using neural signals to inform next-generation search systems.

Research Trajectory

Pollick's research bridges the gap between human cognition and information retrieval (IR) systems by examining the neural correlates of search behaviors. His early collaborative work investigated the temporal dynamics of how users assess search results. As a co-author on When Relevance Judgement is Happening?: An EEG-based Study (2015), he contributed to an electroencephalography (EEG) study analyzing brain signals during the first 800 milliseconds of a relevance assessment process. This research sought to establish a better operational definition of relevance by observing precisely when it physically occurs in the brain, aligning with the broader research direction of Decoding Cognitive States from Distributed Neural Patterns.

Following this, Pollick took a senior role in exploring the foundational concept of "information need"—the often visceral and underspecified state that prompts a user to search. In Understanding Information Need: An fMRI Study (2016), he and his colleagues utilized functional magnetic resonance imaging (fMRI) to investigate how an information need emerges and manifests physically in the brain. Because users are sometimes unaware of their own information needs at the earliest stages, traditional behavioral studies struggle to measure this phenomenon. By mapping these cognitive states directly, this work contributes to the broader area of Resolving Query Underspecification through Interactive Clarification.

Building on the ability to observe the neural signatures of information needs, Pollick's subsequent senior-authored work shifted toward predictive applications. In Towards Predicting a Realisation of an Information Need based on Brain Signals (2019), the research team demonstrated that monitoring brain activity can lead to accurate predictions of when an information need is realized. Conducted via a question-answering task, this study laid the groundwork for proactive IR engines capable of detecting and serving user needs without requiring explicit queries. This trajectory reflects a sustained interest in applying advanced neuroimaging techniques to solve fundamental challenges in human-computer interaction, touching upon themes such as Decoding Neural Representations with Searchlight Mapping.

Current Directions

The most recent papers in the evidence, dating to 2019, indicate a focus on the proactive detection of cognitive states to drive next-generation search systems. This work investigates how continuous monitoring of neural signals can identify the exact moment a user realizes an information need. By moving from descriptive mapping of brain activity to predictive modeling, this research aims to enable systems that anticipate user intent before a formal query is ever articulated. This direction intersects with broader methodological challenges in the field, including the research area of Statistical Generalizability in Neuroimaging, as the development of reliable, predictive brain-computer interfaces requires robust models that can generalize across different subjects and cognitive tasks.

Profile updated

See something wrong? Report a mistake