Using Wikipedia to learn semantic feature representations of concrete concepts in neuroimaging experiments

Abstract

In this paper we show that a corpus of a few thousand Wikipedia articles about concrete or visualizable concepts can be used to produce a low-dimensional semantic feature representation of those concepts. The purpose of such a representation is to serve as a model of the mental context of a subject during functional magnetic resonance imaging (fMRI) experiments. A recent study [19] showed that it was possible to predict fMRI data acquired while subjects thought about a concrete concept, given a representation of those concepts in terms of semantic features obtained with human supervision. We use topic models on our corpus to learn semantic features from text in an unsupervised manner, and show that those features can outperform those in [19] in demanding 12-way and 60-way classification tasks. We also show that these features can be used to uncover similarity relations in brain activation for different concepts which parallel those relations in behavioral data from human subjects.

Speaking, Seeing, Understanding: Correlating semantic models with conceptual representation in the brain

“…Since then, various studies have shown that distributional semantic models encode and are able to predict neural activation patterns associated with concepts (Devereux et al, 2010;Murphy et al, 2012;Pereira et al, 2013). Devereux et al (2010) build on the work of Mitchell et al (2008) and show that automatically acquired feature-norm like semantic representations can make equally powerful predictions about brain activity associated with the presentation of words.…”

“…Devereux et al (2010) build on the work of Mitchell et al (2008) and show that automatically acquired feature-norm like semantic representations can make equally powerful predictions about brain activity associated with the presentation of words. Pereira et al (2013) use semantic features learnt from topic models on Wikipedia to predict neural activation patterns for unseen concepts.…”

“…Mitchell et al (2008) were the first to demonstrate that distributional semantic models encode some of the patterns found in the fMRI data. Other researchers followed in their steps, evaluating traditional count-based distributional models (Devereux et al, 2010;Murphy et al, 2012), topic model-based semantic features (Pereira et al, 2013), psycholinguistic and behavioural features (Palatucci et al, 2009;Chang et al, 2010;Fernandino et al, 2015) and visual representations (Anderson et al, 2013(Anderson et al, , 2017. While all of these studies report correlation between the investigated semantic models and patterns found in the brain imaging data, their focus on individual models and the use of different datasets and prediction methods make their results difficult to compare and to integrate into a coherent evaluation landscape.…”

Automated Generation of Fill-in-the-Blanks-Type Quizzes Using Wikipedia

“…The purpose of such a representation is to serve as a model of the mental context of a subject during functional magnetic resonance imaging (fMRI) experiments [8]. They compile words in classical lists corresponding to concepts that were deemed concrete or imageable, and related articles which are linked to from these article titles in Wikipedia by hand.…”

Seeing is Worse than Believing: Reading People’s Minds Better than Computer-Vision Methods Recognize Actions

“…5].) Pereira et al [20] reanalyze the preceding data in the context of a prior from Wikipedia and achieve a mean accuracy of 13.2% on a one-outof-12 classification task and 1.94% on a one-out-of-60 classification task. Hanson & Halchenko [9] perform classification on still images of two object classes: faces and houses, and achieve an accuracy above 93% on a one-out-of-two classification task.…”