An architecture for skill assessment in serious games based on Event Sequence Analysis
Juan Antonio Caballero-Hernández, Manuel Palomo‐Duarte, Juan Manuel Dodero · 2017
Serious games are video games with educational purposes. Players interact in many points during a gameplay. These interactions can be registered, producing data sets with sequence of events which provide relevant information about player's skills. Unfortunately, traditional skill assessment methods present limitations to carry out a detailed analysis of large data sets. Sequence Analysis is a group of techniques which allow to analyze data sets consisting of sequence of events. These techniques have been successfully implemented in different fields, and we consider that they can help overcome these limitations. In this paper, we propose an architecture of skill assessment in learning experiences based on serious games using a set of Sequence Analysis techniques known as Process Mining. First, several in-game events are stored in a log. These events are produced by player's interactions with Key Performance Indicators included in the game. Second, event log is used as input for a Process Mining tool. Discovery process is executed and a behavioural model is provided. Third, an assessment metric must be carried out over the model. Finally, a synthetic experiment is conducted and promising results are obtained.