Back to top

Collaborative learning video indexing (COLVI)

Last modified Dec 21, 2016


For the last decade digital videos have become ubiquitous, people are watching and uploading videos on YouTube, creating video podcasts and screencasts, filming events in real time using their smartphones or tablets. Use of digital videos in education can be listed as one of the video prevalence examples as well. Since educational process has been extended by possibility to get qualitative knowledge from online courses, online academies, online tutorials and online demos, videos start to play one of the most important roles in such online education environments. At the same time the increased amount of the online courses has enhanced the consumers’ expectations. Quality of video content, structure, correspondence of the offered material to the level of students’ knowledge and many other factors are influencing consumers’ preferences. To meet consumers’ expectations and improve educational environment, different metrics (ratings, viewing time, “intensity”, feedback comments) can be collected and analyzed.

Current project goal and challenges

To narrow the scope of the initial work and experiment with concrete problem example, the development of the web application, that could help Bachelor student at the faculty of informatics at TU Munich easily find and watch short, useful explanatory video clips for any question related to the topic of his or her courses, will be one of the main project goals.

Two main challenges are identified:

  • The quality of video content is difficult to access 
  • The lack of meta-data, navigation aids and content descriptors lead to the poor video content findability.

Research questions

The following research questions are addressed:

  • How to identify the possible questions arising in the courses offered for Bachelor Students at TU Munich?
  • How to identify explanatory video clips for these topics?
  • How to obtain indicators on the quality of the video clips from the viewpoint of the target audience?
  • How to motivate the community members to upload, segment, annotate and rate?


[Ma14] Florian Matthes, Klym Shumaiev. Collaborative Annotation of Recorded Teaching Video Sessions6th International on Conference Computer Supported Education (CSEDU 2014), p. 576-581, Scitepress, Barcelona, Spain, 2014