< Back to previous page


Medial temporal representation of word meaning embedded in natural language processing

Sentence comprehension deficits can arise despite preserved single word comprehension due to a deficit of integrating information into an overall meaning representation. In this proposal, I will leverage the rapid advances in Natural Language Processing (NLP) in computational sciences to investigate the neurobiology of natural language comprehension. Five fMRI and two eye tracking experiments will be conducted to examine how the brain produces different conceptual representations for the same concept in a context-relevant manner during connected language, with a special focus on the role of anterior and medial temporal cortex. Investigation of comprehension of connected speech will be gradually built up across experiments: I will compare the representation of meaning of single words to that of single words embedded in a sentence presented with Rapid Serial Visual Presentation (RSVP) (Exp1) - in written and auditory modality (Exp2). Next, I will compare paragraphs presented with RSVP to immediate presentation of the written paragraph (Exp3). Finally, I will test the comprehension of general knowledge, episodic schema and personal experience paragraphs in healthy control (Exp4) and aphasia patients (Exp5). Data will be analyzed by means of a voxel-wise modelling approach and Representational Similarity Analysis. Moreover, paragraphs will be converted in paragraphs vectors so that a computational investigation of their similarity will be addressed.

Date:1 Oct 2020  →  Today
Keywords:connected speech, natural language processing, semantics
Disciplines:Cognitive neuroscience, Data visualisation and imaging , Numerical computation, Neuroanatomy