Call for papers: Mining Performance Patterns in Elite Sports

Calling on researchers in the different fields of Machine Learning, Statistics, Data Mining, and Sport Science to submit papers for the workshop in the 13th IEEE International Conference on Data Mining series (ICDM). The workshop will be held in Dallas, Texas 7 December, 2013.

Papers will discuss related topics of the applications, current challenges, and possibly the future of using advanced data analytics techniques in knowledge discovery and knowledge generation in elite sports.


Topics of interest include but are not limited to:

  • Overview of data mining in elite sports
  • Frameworks and challenges to bring together data mining and sport science
  • Application of existing data mining and machine learning techniques in sports data analysis
  • Challenges/benefits of using major and current data mining and machine learning techniques in sports. For example:
    • Supervised learners
    • Unsupervised learners
    • Probabilistic learners/reasoners
    • Rule miners
    • Statistical relationship estimators
  • Adapting data mining and machine learning techniques for sports data analysis
  • Selecting specific data mining techniques for problem types in sports
  • Outlier modelling in sports data
  • Data staging (collection and pre-processing) for sports data mining
  • Real-time data mining-based decision support tools in sports
  • Visualization of sports performance patterns
  • Video and image mining in sports
  • Text mining (and its applications) in elite sports
  • Preparation and interpretation of sports data mining results analysis.

Key dates

Paper submission: 18 August 2013

Reviews start: 18 August 2013

Reviews due: 20 September 2013

Notification to authors: 24 September 2013

Camera ready submissions: 15 October 2013

Workshop: 7 December 2013 (Dallas, Texas)

Topic in depth

Mining Performance Patterns in Elite Sports (MPPES) is the IEEE ICDM workshop on using advanced data analytics techniques for decision making in the elite sports domain. Sports performance analysis is a means to create and analyse a valid record of athlete performances by using systematic observations. Many researchers are keenly interested in the intersection of the two domains of Information Technology and Sports Science. It has gained importance in the last decade due to advances in Information Technology and Digital Photography. Machine Learning and Data Mining are currently used to analyse a variety of sports data.

Data from international competitions such as World Championships and Olympic Games provide opportunities for mining elite sports performance patterns.

Most well-established data mining and machine learning techniques have been applied in modelling and mining sports data. Unsupervised learning (e.g. clustering), supervised learning (e.g. classification), relationship estimation (e.g. regression analysis), and rule mining (i.e. sequential patterns and association rules) techniques have widely been applied to a number of decision problems in the elite sports domain. Such ad hoc and non-systematic single problem-oriented efforts have never been fully integrated in to generic problem type-centred solutions and structures.

The emerging field of Elite Sports Data Mining still lacks a covering framework of techniques and guidelines for the different decision problems in this domain. Although elite sports data inherit some unique characteristics (for example outlier performances often result in long-standing world records and are thus very interesting), specific data mining techniques and solutions have never been adapted in such a way to become sports domain-compatible. The latter issue is realised especially in the two main steps of sports data mining, namely, data staging (collection and pre-processing) and analysis.

The elite sports domain involves a number of rules, regulations, tactics, strategies, performance measures, conditions, abilities, and performance evaluation criteria related to each specific competition. These aspects, as well as decision problem types (for example performance pattern discovery and performance prediction), require careful consideration when applying any data staging techniques.

Sports decision problems and problem types also play an important role in deciding what data mining or machine learning technique is suitable to be utilized in the analysis step.

It's important to understand which criteria to consider when selecting or adapting a specific data analytics technique for specific decision problem types in the elite sports. This will provide the ability to optimise utilisation or development of efficient and effective techniques well-suited in this domain.

Submission guidelines

We invite regular paper submissions, work-in-progress, demos, and position papers. All  papers must follow the IEEE ICDM submission guidelines.

Submit your paper by 18 August 2013 .

Regular papers can be up to 8 pages in length; short papers, such as position and work-in-progress papers, no more than 6 pages; two additional pages can be purchased for $125 per page. All papers will be reviewed by the Program Committee on the basis of technical quality, relevance to workshop topics, originality, significance, and clarity.


Information about how to register will be available from the IEEE ICDM 2013 website once registrations are open. We hope to meet you all in Dallas, Texas during the workshop in December 2013.

Workshop organisers

Dr Bahadorreza Ofoghi, National ICT Australia and the University of Melbourne, Australia

Prof. John Zeleznikow, Victoria University, Australia

Dr Clare MacMahon, Swinburne University, Australia

Dr Dan Dwyer, Deakin University, Australia

Contact us

Further questions regarding submissions and participation in the workshop, please contact one of the organisers (include IEEE ICDM Workshop: MPPES in the subject line):

Dr. Bahadorreza Ofoghi: [email protected]

Prof. John Zeleznikow: [email protected]

See all news