Master Thesis: Study on the Application of Graph Algorithms and Social Network Analysis for Serious Gaming (deutsch/english)

(Deutsch oder Englisch)

Master thesis on learning analytics with graph algorithms for serious gaming. The research question is: how to apply graph algorithms to serious games and standardized usage data to explore learning analytics insights, and to analyse usage patterns with social network analysis strategies.
The application domain of this applied research work is aerial image exploitation for which the Fraunhofer IOSB is developing intelligent assistance systems, one being an adaptive serious game. Key aspects of this work are the integration of a prototype to collect usage data, a technical study on graph algorithms and an evaluation with real usage data.


Key Aspects & Challenges

  • Aplication of social network analysis methods and graph algorithms (from e.g. Neo4j) to serious games learning analytics.
  • Integration  of  a  running  prototype  application  including  the  serious  game ”Lost Earth” from the Fraunhofer IOSB.
  • Collection of usage data to evaluate graph algorithms application, starting with synthetic data for the conceptual verification, later with real user interaction data for an evaluation study.
  • Discussion of advantages and limitations of the approach.
  • Usage  of  modern  e-learning  interoperability  standards  and  methodologies, e.g., Experience API (xAPI) and use of ”E-Learning A.I.” (ELAI).



  • Conducting a systematic literature research on the current state of the art in science and technology.
  • Formal conceptualization and realization of an integrated prototype application which includes the serious game, graph database and analytics dashboard.
  • Generation of synthetic usage data (from Lost Earth or ElaiSim usage data generator) and storage in Learning Record Store (LRS) and Neo4j database.
  • Conducting a technical study on the application of relevant graph algorithms to the collected usage data. Relevant algorithms comprise social network analysis methods, e.g., detection of important users or learning objects with centrality or neighborhood graph algorithms.
  • Discussing the advantages and limitations of the approach. Clearly stating the benefits of the gaming/learning analytics for the various user roles, i.e., for the user/learner and for the teacher/tutor.
  • Conducting a study with real users playing the game to verify the approach with real usage data. Discussion of the results.


  • Interest and motivation.
  • Computer science with knowledge in graph structures and optimally graph algorithms.
  • Best: knowledge of xAPI, triple structures, graph databases like Neo4j, social network analysis.

Please send your application to Alexander Streicher.