Maps of meaning in the brain and their data management

Grasping the meaning of a word implies the activation of an incredible richness of information, which spans from vision (e.g., a banana is yellow), to other perceptual modalities (it has a specific smell), function (it is edible), categorical structure (it is a fruit), general encyclopedic knowledge (it grows in tropical areas), and much more. How the brain masters this extremely complex network is still largely unknown. One of the reasons for this is that it's experimentally difficult to gauge this network at scale; most studies in Neuroscience and Cognitive Psychology are limited to a few tenths of words/concepts, or contrast broad categories (e.g., tools vs. animals). In this seminar, I'll illustrate our attempts at breaking this boundary and provide a large-scale characterization of the human semantic network, using Electroencephalography (EEG) and computational models (e.g., vector semantics). Our initial analyses show that there's a latent common structure in different kind of grounded information (e.g., language vs. vision), which the brain captures (of course) and deploys at different times in different brain areas.