CS Special Seminar: Tuukka Ruotsalo "Cognitive Interfacing with Data"
Cognitive Interfacing with Data
Thursday, 19 August at 10:30
via Zoom: request the link by email [email protected]
Note! the link will be sent to the CS staff.
Abstract: The ability to understand human needs and communicate them with computers is one of the hallmarks of human-centered computing, data management, and information retrieval. What if we could just naturally react to information that we encounter and computing systems could learn our preferences, optimal representations of data, and assist us to store, model, retrieve and generate information by sensing our thoughts and reactions? The present approaches to utilize information from user interactions in such tasks rely on simple observations of explicit interactive behavior. Even the current physiological and brain-computer interfaces (BCI) still require their users to actively think an artificial pattern that corresponds to a command, and the computer can then interpret this pattern to execute the command. For example, a classic BCI speller requires its user to actively attend to a certain letter on the screen and tests all letters presented on the screen to detect and spell the one the user is attending to.
In this talk, I will present cognitive interfacing with data -- a new line of research in which passive or natural human reactions in response to perception of data are interpreted via brain-computer interfacing to guide machine-learned models of the data. Cognitive interfacing mitigates the need of explicit man-machine interaction with computing in various data-intensive applications. I will demonstrate applications of such cognitive interfacing in information retrieval, recommender systems, crowdsourcing, and interactive generative modeling.
Bio: Tuukka Ruotsalo is an Associate Professor of Computer Science at the University of Copenhagen, Denmark and holds an Academy of Finland Research Fellowship at the University of Helsinki, Finland. Tuukka leads the cognitive computing group focusing on the use of machine learning to interpret signals originating from human cognition, and the use of those signals to augment machine learning, data management, information retrieval and man-machine interaction. His research is published in the most prestigious forums of data management and information retrieval, such as SIGIR, CIKM, TOIS, JASIST; human-computer interaction, such as CHI, UIST, and IUI; as well as general science outlets, such as Scientific Reports and the Communications of the ACM. Earlier in his career, Tuukka has held academic posts at UC Berkeley and Harvard University and is a recipient of several awards, including the World Summit Award of the United Nations.
More information: https://www.cs.helsinki.fi/group/intercom/