Information Technology Reference
In-Depth Information
Fig. 12.4
Role of process modeling for analyzing event logs (modified after van der Aalst &
Günther, 2007).
•
Discovery—No a priori model exists. Based on a event log a model is con-
structed;
•
Conformance—An a priori model exists. Event logs are used to determine the
extent to which the enacted collaboration corresponds to the model.
•
Extension—an a priori model exists. The goal is not to test but to extend
the model, for instance, with performance data (e.g., durations of activities).
Extended models can then be used for example to optimize the process (Van
der Aalst & Günther, 2007).
To the extent that the algorithms can be sufficiently tuned to a specific area of
individual or collaborative learning to deliver meaningful results, they could be
run without human intervention and hence realize a (quasi-)continuous analysis of
(quasi-)continuous data streams.
Collaborative Analysis of Video Recordings
Video as a data format to analyze and learn from is highly valued in education,
both by practitioners as well as by researchers (Goldman, Pea, Barron, & Sharon,
2007). Like no other medium, video recordings (including audio) can store con-
textual information, which is particularly important for recording practices, such as
classroom practices, which are not predominantly taking place in a textual medium.
Since video recordings can store so much (but certainly not everything) of the con-
text, they also allow the researcher to defer decisions as to which information to
attend do in detail to later (thus contributing in no small measure the
data deluge
typical for qualitative research).
However, at this stage the analysis of video data is a very work-intensive practice.
Depending on the questions asked and the analysis method employed, the relation
