Alfred Kobsa

Selected Research Projects

Cooperation Awareness and Privacy

This project aims at analyzing privacy and awareness needs of [distributed] workgroups, and at supporting workgroups in balancing the demand for awareness information with individual privacy preferences.


Alfred Kobsa (Faculty)
Sameer Patil (Ph.D. student)
Doree Seligmann and Ajita John, Avaya Research

Duration: 2002-2010  
Funding: NSF  ("mid-size" ITR grant)
NSF (for collaboration with the European PRIME project)
NSF (large HCC grant)

J25 J26, J27, C34, C36, C37, C42, C46, C52, C53, B17, B23, W21, W27, O15

Privacy as a Design Requirement for Personalized Systems

Privacy demands of Internet users and international (and future national) privacy legislation have an impact on the collection of personal data in web-based systems. This project studies specifically the impacts on "personalized" web-based systems, which cater their interaction to each individual user, collect considerable amounts of personal data for this purpose, and "lay them in stock" for possible future adaptation. The project analyzes and documents these privacy requirements and provides solutions for the software architecture and the user interface of personalized systems to cater to privacy constraints of each individual user and their jurisdictions.


Alfred Kobsa (Faculty)
Yang Wang (Ph.D. student)
Max Teltzrow (visiting Ph.D. student)
Zhaoqi "Stella" Chen (Ph.D. student)

Duration: 2001 - 2010
Funding: CRITO (2001-03), NSF (since July 2003), Humboldt Foundation (since Oct. 2004, for collaboration with Humboldt University), Google Research Award

J17, J23, C25, C37, C38, C43, C47, B15, B16, B20, B21, W15, W16, W17, W18, W19 , W22, W25, O32

Catalog of requirements imposed by international privacy laws

Privacy through Pseudonymity in User-Adaptive Systems

Adaptive systems are generally better able to cater to users the more data their user modeling systems collect and process about them.  This project analyzes security requirements to guarantee privacy in user-adaptive systems and explores ways to keep users anonymous whilst fully preserving personalized interaction with them. User anonymization in personalized systems goes beyond current models in that not only users must remain anonymous but also the user modeling system that maintains their personal data. Moreover, users' trust in anonymity can be expected to lead to more extensive and frank interaction, hence to more and better data about the user, and thus to better personalization. A reference model for pseudonymous and secure user modeling has been developed and implemented that meets many of the proposed requirements.


Alfred Kobsa (Faculty)
Jörg Schreck (Ph.D. student)

Duration: Oct. 1996  - July 2001  
Funding: GMD (Fraunhofer)  
Results: J20, C16
Jörg Schreck (2003): Security and Privacy in User Modeling. Dordrecht, Netherlands: Kluwer Academic  Publishers.

Success Factors of Information Visualization Systems

Research in information visualization aims at leveraging human perceptual abilities for data analysis, by presenting information not as numbers and text but in visual form. Effective visual displays must therefore be designed in such a way that users can easily obtain an overview of the data, spot outliers, patterns and correlations, and see the changes that will occur if some parameters become altered. Visualizations need to make such significant facts salient for human perception. This identifies "success factors" of visualization systems for multivariate data, by conducting experiments both with analysts working individually on a computer screen, and with groups of analysts collaborating in front of a large-sized display. 


Alfred Kobsa (Faculty)
Victor Gonzales (Ph.D. student)
Several undergraduate students

Duration: Since Jan. 2001  
Funding: CRITO  

J21, C24, C27 , C29, C30, C31, C32, C35, O17

User Modeling as an Open Learning Process (LaboUr)

Generic user modeling systems so far pursued a strongly knowledge-based approach. Heuristics regarding assumptions that can be made about users when certain observations were made about them had to be empirically found beforehand and expressed in knowledge representation mechanisms. The project LaboUr developed a generic prototype that incorporated user modeling as an open learning process. "Open" thereby means that the user modeling system can process any kind of assumption about the user, and can communicate with any kind of source that provides information about the user. Learning "process" refers to the continuous  incremental processing, abstraction and revision of assumptions about the user based on the observed user behavior.


Alfred Kobsa (Faculty)
Wolfgang Pohl (Post-Doc.)
Ingo Schwab (Ph.D. student)
Ramin Yasdi (Researcher)

Duration: Aug. 1997 – Nov. 2000  
Funding: German Research Foundation  

J18, W12
Schwab I., Pohl W. and Koychev, I. (2000). Learning to Recommend from Positive Evidence, Proceedings of Intelligent User Interfaces 2000, ACM Press, 241-247.
Pohl W. (1997). LaboUr - Machine Learning for User Modeling. Proc. of the Seventh International Conference on Human-Computer Interaction. Amsterdam: Elsevier.

Adaptive and Adaptable Interactions for Multimedia Telecommunications (AVANTI)

The World Wide Web is going to be the leading online information medium for some years to come and will most likely become the standard gateway for citizens to the “information highway”. Already today, visitors of web sites are generally heterogeneous and have different needs, and this is likely to increase in the future. The aim of the AVANTI project in which 12 organizations collaborate is to cater hypermedia information to these individual needs by adapting the content and the presentation of web pages to each individual user. The special needs of elderly and disabled users are also partly considered. A model of the characteristics of user groups, individual users and usage environments, and a domain model are exploited in the adaptation process. One aim of this research is to verify that adaptation and user modeling techniques that were hitherto mostly used for catering interactive software systems to able-bodied users also prove useful for adaptation to users with special needs. Another original aspect is the development of a network-wide user modeling server that can concurrently accommodate the user modeling needs of several applications and several instances of an application within a distributed computing environment.


Alfred Kobsa (Faculty)
Josef Fink (Ph.D. student)
Andreas Nill (Ph.D. student)
Igor Jaceniak (Researcher)
Armin Hüttenhain (Researcher)
Gabriele Nordbrock (Researcher)
Several undergraduate students

Duration: Aug. 1995 – Nov. 1999  
Funding: European Commission (ACTS Programme)
Collaborative research of 12 institutions
Results: J12, C14, C15, C16, C17, C18, C19, C20, C22, B12, N10, O11

A User Modeling Shell System (BGP-MS)

This project developed BGP-MS, a user modeling shell system that can assist interactive software systems in adapting to their current users by taking the users’ presumed knowledge, beliefs, and goals into account. It offers applications several methods for communicating observations concerning the user to BGPMS, and for obtaining information on currently held assumptions about the user from BGP-MS. It provides a choice of two integrated formalisms for representing beliefs and goals, and includes several types of inferences for drawing additional assumptions based on an initial interview, observed user actions, and stereotypical knowledge about pre-defined user subgroups. BGP-MS is a customizable software system that is independent from applications, operates concurrently with them, and interacts with them through inter-process communication. For tailoring BGP-MS to a specific application domain, the developer must select those components of BGP-MS that are needed in this domain and fill them with relevant domain-dependent user modeling knowledge.


Alfred Kobsa (Faculty)
Wolfgang Pohl (Ph.D. student)
Jörg Höhle (Ph.D. student)
Several undergraduate students

Duration: Feb. 1992 – Aug. 1997  
Funding: German Research Foundation  

J10, C12, C13, C21, C23, C26, W6, W8, W9, W10
Server software:
Draft standard: O10
Wolfgang Pohl (1997): Logic-Based Representation and Reasoning for User Modeling Shell Systems. Berlin: Academic Publishing Corporation – infix.

A Natural-Language Access System to Expert Systems (XTRA)

The XTRA access system to expert systems aims at rendering the interaction with expert systems easier for inexperienced users. XTRA communicates with the user in a natural language (German), extracts data relevant to the expert system from his/her natural-language input, answers user queries as to terminology, and provides user-accommodated natural-language verbalizations of results and explanations provided by the expert system. A number of novel artificial intelligence techniques have been emplozed in the development of the system, including the combination of natural-language user input with user gestures on the terminal screen, referent identification with the aid of four different knowledge sources, simultaneous communication of the access system with the user and the expert system, fusion of two complementary knowledge bases into a single one, and the design of a natural-language generation component which allows for a controlled interaction between the "what-to-say" and the "how-to-say" parts to yield a more natural output. XTRA is being deveoped independenly of any specific expert system. In its first application, the access to an expert system in the income tax domain is being realized.


Wolfgang Wahlster (Faculty)
Alfred Kobsa (Co-director)
Roman Jansen-Winkeln (Ph.D. student)
Jürgen Allgaier (Ph.D. student)
Carola Reddig (Ph.D. student)
Dagmar Schmauks (Researcher)
Several undergraduate students

Duration: May 1985 – Sept. 1991  
Funding: German Research Foundation  
Results: J5, J6, J7, J8, J9, C7, C8, C9, C10, C11, B8, B9, W1, W2, W3, W4, W5, N6, N7, N8, N9, O6