Year: 2011
Type: dissertation
Abstract: Recent advances in human brain imaging techniques have shown that it is possible to correctly identify thoughts associated with specific categorises of conceptual knowledge. A semantic space is created through word clustering from a corpus and cross-referenced with words read during a natural reading task while at the same time recording EEG. These results show that it is possible to correctly classify a words category 25% of the time (out of 12 possible categories) based on neural activity within the N400 component with the help of a artifical neural network. Results also visualises the ANN’s black-box behaviour in which electrodes it deem most contributing when classifying. These results contribute to the discussion on how the semantic memory is stored in the brain; modularly or connectionisticly, and also how far it is possible to subcategories thoughts and still correctly identify them with brain imagining techniques.
Author: Johan Emilsson
Institution: Lund University
Language: English
Subfield: Cognitive Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W1782192584