Neural responses to abstract and linguistic stimuli with variable recognition latency

Abstract

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 tar- get 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. Dis- criminative information could be extracted better if the EEG data were aligned to the response than to the stimulus-onset.