Since 2013, the German Association for Sports Science (dvs) has awarded an annual PostDoc Lecture to two young scientists with doctorates, giving them a forum within the framework of the Sports Science University Day to report on their research to a larger professional audience.
"With the awarding of the PostDoc Lecture, the dvs has an instrument to support young scientists," said Prof. Dr. Ansgar Schwirtz, dvs President and Head of the Associate Professorship of Biomechanics in Sport. "This lecture is the highest prize we award, in this respect we are also very proud of the prize winners."
Among nine applications in two categories, the three-member jury selected this year, in addition to Dr. Birgit Braumüller (German Sport University Cologne), Dr. Otto Kolbinger, who has been working as a Research Associate at the Chair of Performance Analysis and Sports Informatics since 2013. He prevailed in the Natural & Behavioral Science topic area with his abstract on "Text Mining in Training Science and Competitive Sports."
"Since there are not many awards for postdocs, this award is a nice confirmation for me that the topic of text mining is still very rarely used, but is definitely of interest to many scientists," explains Dr. Kolbinger. "In this respect, it was of course great to receive such a stage at the dvs University Day and to present text mining in the context of a method talk as well as to show what potential still exists here."
The text mining method, similar to data mining, automatically derives knowledge from large text datasets. This approach is mainly used to identify themes or mood (sentiment) of texts and text segments or to predict behavior or events. In the field of sports science, the method has so far been used mainly in sports sociology. Therefore, Dr. Kolbinger presented in his lecture how text mining can also be used in performance analysis and in competitive sports in general. As application examples he presented on the one hand the evaluation of rule changes and on the other hand the field of talent scouting.
For example, the Video Assistant Referee (VAR) in soccer was used to show how social media posts can be used to examine the effects of rule changes on the opinion of stakeholders. Text mining can also be applied in talent scouting, for example, when sentiment analysis using lexicon-based categorization is used to evaluate scouting reports in ice hockey.
"Text mining is an unusual approach for processing social media posts, for example, in terms of content analysis on the topic of sentiment analysis," says Prof. Dr. Martin Lames, head of the Chair of Performance Analysis and Sports Informatics. "The method has so far been used more in sports management and the field of image research. In performance analysis, on the other hand, the method has not yet been so widely used. Based on the outstanding award for Otto Kolbinger, however, you can see the relevance here for our research area."
To the homepage of the Chair of Performance Analysis and Sports Informatics
Contact:
Prof. Dr. Martin Lames
Chair of Performance Analysis and Sports Informatics
Georg-Brauchle-Ring 60/62
80992 München
phone: 089 289 24500
e-mail: Martin.Lames(at)tum.de
Dr. Otto Kolbinger
Chair of Performance Analysis and Sports Informatics
Georg-Brauchle-Ring 60/62
80992 München
phone: 089 289 24502
e-mail: Otto.Kolbinger(at)tum.de
Prof. Dr. Ansgar Schwirtz
Associate Professorship of Biomechanics in Sports
dvs President
Georg-Brauchle-Ring 60/62
80992 München
phone: 089 289 24
e-mail: Ansgar.Schwirtz(at)tum.de
Text: Romy Schwaiger
Photos: dvs