Fakultät Wirtschaftsinformatik und Angewandte Informatik

Angewandte Informatik / Kognitive Systeme

Bachelor-Projekt Kognitive Systeme/Project Cognitive Systems

General Information

  • For a general course description please read the corresponding pages from the WIAI module guide.
  • Participants should sign up for the course in the virtual campus at least a week before the semester starts or write an EMail to the lecturer.
  • You find administrative information at UnivIS.
  • Suggested topics can be found here.

Proposed Topics

  • We have a shared pool of topics for bachelor- and master-level projects as well as for bachelor- and master theses, which are described in the VC-Course "KogSys: Allgemeine Informationen": VC-Course.
  • Selection of research domains addressed in the projects: Inductive programming, applications of machine learning to image classification (e.g., pain faces, plant growth), grammar learning, Human-robot interaction with NAO, Cognitive models of analogy and inductive learning.
  • We strongly recommend students interested in writing a thesis in our group to participate in a project before!
  • Topics will be assigned based in interest and methodological background of the students.
  • A topic is usually worked on jointly by two to four students.
  • A list of past and current project topics is given below.

Current Projects

Past Projects

WS 2013/2014

Topic: Classifying Car Shapes with High-Level Features

Description: In this project characteristic features of different car models of the vendor Audi are identified. Then recognition techniques for these features are explored and evaluated. (Supervisor: Michael Siebers)

Students: Florian Muth (MA AI), Sebastian Ulschmid (BA AI)

Report: [pdf]


SS 2013

Topic: A cognitive model of aesthetical preferences based on dynamic prototypes

Description: Based on a masters project in the previous semester, new data will be obtained which allow for adaptation in both dimensions of each given feature. Furthermore, a new model which allows incremental shift will be tested with the new data. (Supervisor: Michael Siebers)

Students: Tina Kämmerer, Ioulia Kalpakoula, Michael Kleber 

WS 2012/2013

Topic: NAO project

Description: Exploring programming NAO with a simple task: picking up a red ball, placing it and kicking it into a goal. Providing a programmers manual for NAO. (Supervisor: Ute Schmid)

Students: Jaime A. Delgado Granados, Fabian Selig, Norman Steinmeier, Daniel Winterstein

Report: [pdf] Manual: [pdf] SourcesChoreosMovies


SS 2012

Topic: Induction of Number Series

Description: Based on an analysys of complexity of different types of number series problems, we will conduct an empirical study to test these domains. Furthermore, a computer model about solving number series will be realized. (Supervisor: Ute Schmid)

Students: Roland Grünwald, Elke Heidel, Alexander Strätz, Michael Sünkel, Robert Terbach

Report: [pdf]

Topic: Implementing a framework for Emotion Analysis from video streams of facial expressions

Description: A video analysis framework is the next milestone in the ongoing facial expression analysis study (PainFace, in colaboration with Biological Psychology Group). The aim of this framework is to provide components that can be combined to a software which analyses video streams and presents the results. (Supervisor: Michael Siebers)

Students: Klaus Schneider

Report: [pdf]

WS 2011/2012

Topic: Implicit to Explicit Learning -- Combining ACT-R problem solving with inductive programming

Description: Problem solving in ACT-R relies on a fixed interpreter strategy. Changes in problem solving strategies with experience are typically modeled by modifications of strength values of rules. Th inductive programming system Igor, on the other hand, learns new problem solving stragegies by generalization over problem solving traces. The strategies are represented as sets of productive (i.e. recursive) rules. We want to introduce this concept into ACT-R. In a first step, we will model productive rule sets as chunks in declarative memory together with a set of general production rules which realise application of such declarative chunks to given problems. We will investigate the advantage of such a strategy for classical problems such as Tower of Hanoi. (Supervisor: Ute Schmid)

Students: Tina Kämmerer, Ioulia Kalpakoula, Michael Kleber


SS 2011

Topic: Implicit and explicit learning of artificial grammars from letter strings vs visual patterns vs visual-motor patterns

Description: We plan and conduct a psychological experiment for artificial grammar learning.  We are interested in the effect of stimulus presentation (letter strings vs. 2D grid patterns vs. 2D grid patterns with motor interaction) on the quality and explicitness of learning.

Students: Peter Grossmann (Master PJ), Jacqueline Hofmann, Tobias Kaiser, Simone Schineller, Johannes Folger, Dominik Seuss

Report: [pdf] Slides: [pdf]

WS 2010/2011

Topic: Constructing sensumotoric maps from interaction

Description: Using a Legomindstorm robot which explores its environment, a sensumotor map is learned from interactive experience.

Student: Mark Wernsdorfer (BA AI)

Report: [pdf]

SS 2010

Topic: Building a Mobile Eye Tracker

Description: We realize a mobile eye tracker and provide a software tool for analyzing eye tracking data. The eye tracker is practically tested in the context of an empirical study in the domain of problem solving.

Students: Thomas Heinz with Michael Albert (MA AI) and Mike Imhof (Psychology)

Report: [pdf]

WS 08/09

Topic: Pain assesment from facial expressions

Description: Based on a sample of meshes gained from facial images with pain expressions it will be explored whether and in what degree individual classifiers will be more precise in pain recognition than classifiers learned over a sample of persons.

Student: Michael Sieber

Talk: [pdf]

SS 08

Topic: Computersimulationen als Instrument zur Erforschung kognitiver Prozesse und gemeinsamen Lageverständnis bei verteilt arbeitenenden Einsatzteams (In Kooperation mit Prof. Dr. Harald Schaub und Dr. Frank Detje, IABG mbh)

Description: Die erfolgreiche Koordination von verteilt arbeitenden Einsatzteams -- im Katastrophenschutz, von Krisenstäben und anderen Einsatzkräften -- auch über große räumliche Distanz beeinflusst in hohem Maß den Erfolg solcher Einsätze. Mit zunehmender technischer Vernetzung steigen die kognitiven Anforderungen an die Akteure bei der Koordination der Einsatzkräfte. Ziel des Projektes ist die Entwicklung einer Computersimulation, basierend auf der Feuer-Simulation von Brehmer und Dörner. Eine vernetzte Version, implementiert in Java, soll eine Grundlage liefern für eine empirische Evaluation möglicher kritischer Faktoren bei der Kommunikation verteilt arbeitender Teams. Als theoretische Grundlage dient die aktuelle Forschung zu Situation Awareness, zum verteilten Lösen komplexer Probleme und zum Umgang mit komplexen Situationen.

Students: Thomas Bornschlegel, Florian Muth, Christian Nappert, Marius Raab

Talk: [pdf]

SS 05

Topic: Kann menschliches Konzeptlernen als inkrementelles Entscheidungsbaum-Lernen modelliert werden?

Description: Es soll ein psychologisches Online-Experiment durchgeführt werden, mit dem die Hypothese überprüft werden soll, dass menschlicher Konzepterwerb als inkrementelles Entscheidungsbaum-Lernen modelliert werden kann. Operationalisierung 1: Die Reihenfolge der Trainingsbeispiele beeinflusst die Komplexität des gelernten Baums und damit die Entscheidungszeiten bei nachfolgenden Klassifikationsaufgaben. Operationalisierung 2: Im Entscheidungsbaum werden unter bestimmten Bedingungen nur die relevanten Merkmale repräsentiert, beim exemplarbasierten Ansatz dagegen alle Merkmale. Entsprechend müssten die Klassifikationszeiten im exemplarbasierten Ansatz höher liegen als beim Entscheidungsbaumansatz, wenn im Entscheidungsbaum nur wenige Merkmale Eingang finden.

Students: Phillip Merensky, Michael Räther, Eric Steinkamp


  • Modellierung Menschlichen Konzepterwerbs durch inkrementelles Entscheidungsbaum-Lernen [pdf]
  • Realisierung eines online-Experiments mit Zeiterfassung [pdf]


  • CAL2 und CAL3 aus: Unger, S. und Wysotzki, F.: Lernfähige Klassifizierungssysteme
  • Tom Mitchell (1997). Machine Learning. McGraw Hill. (Kapitel Decision Tree Learning, Instance-Based Learning)
  • Waldmann, M. R. (2002). Kategorisierung und Wissenserwerb. In J. Müsseler & W. Prinz (Hrsg.), Lehrbuch Allgemeine Psychologie (S. 432-491). Heidelberg: Spektrum Verlag.

Links zu online-Experimenten: