Neural Responses to Abstract and Linguistic Stimuli with Variable Recognition Latency
Electroencephalography (EEG) can provide information about which words or items are relevant for a computer user. This implicit information is potentially useful for applications that adapt to the current interest of the individual user. EEG data were used to estimate whether a linguistic or abstract stimulus belonged to a target category that a person was looking for. The complex stimuli went beyond basic symbols commonly used in brain-computer interfacing and required a variable assessment duration or gaze shifts. Accordingly, neural processes related to recognition occurred with a variable latency after stimulus-onset. Decisions involving not only shapes but also semantic linguistic information could be well detected from the EEG data. Discriminative information could be extracted better if the EEG data were aligned to the response than to the stimulus-onset.
KeywordsEEG Single trial classification Physiological computing User relevance estimation
Unable to display preview. Download preview PDF.