Early Stage Researcher at TU Delft

Alisa has a background in congitive science and Human-Machine-Interaction and she is working on the mitigation of cognitive biases through interactive interfaces for recommender systems at TU Delft. She loves gardening and being outdoors.

TU Delft - Technische University of Delft (Netherlands) Department of Software Technology

Department of Software Technology
Technische University of Delf (TU Delft)
Faculty of Electrical Engineering, Mathematics and Computer Science
Van Mourik Broekmanweg 6,
2628 XE, Delft, Netherlands


Interactions to mitigate human biases.

·        Main Supervisor: Dr. Nava Tintarev, Department of Software Technology – Technische University of Delft (TU Delft),  

·         PhD Co-Supervisor: Dr. Mariët Theune, Human Media Interaction Research Group – Universiteit Twente (UTWENTE)

·         Inter-sectoral Secondment Supervisor: Dr. Thijs Westerveld, R&D department – Wizenoze B.V.

PhD research topic


To develop new interaction approaches for recommender systems with the aim to mitigate cognitive biases (e.g., confirmation bias, availability bias, attention heuristic, and backfire effect), which lead to deviation from optimal choices. This will be done through mixed-initiative (automated and human) “conversational” interfaces extending the concept of faceted search while aiding understanding why certain content (even if it is surprising) is being recommended. The interaction will be designed in a way that meaningfully compensates for cognitive bias.

Use cases

Theoretical contributions will be validated in three use cases for recommender systems:

(1) news;

(2) social media; and

(3) educational search.

Expected Results:

New theory and tools for interactions that mitigate human bias:

(1) new interaction approaches that take into account user needs, values, and goals.

(2) Improve explainability of recommender systems by creating a fundamental link between algorithmic approaches and interaction approaches.

(3) Develop new control mechanisms suitable for individual and situational characteristics.

(4) Develop standards for user interfaces and tools for recommender systems.