Tapping neural correlates of the depth of cognitive processing for improving human computer interaction

Abstract

The classical interaction between human and a computer or a machine relies solely on explicit behaviour (input with keyboard, mouse, gestures etc.). In many situations and tasks, the access to implicit information about the user could enhance human-computer interaction (HCI). Recent research has shown a number of examples of how such hidden user states could be extracted from signals of peripheral physiology and of the brain. While these approaches are still premature and not readily available for real application, further exploration seems worthwhile. Here, we present an approach towards monitoring the level of cognitive processing. A special experimental paradigm has been de- signed to detect event-related potentials (ERPs) of brain activity related to cog- nitive processes using tasks in different cognitive domains. Neural correlates indicating different levels of cognitive processing have been singled out and the classifiability was quantified using multivariate decoding methods. The results indicate the feasibility of monitoring the depth of cognitive processing for neu- rotechnological applications in BCI and industrial scenarios.