Research Centre for Data Analytics and Cognition La Trobe University, Bundoora, Victoria 3086, Australia.

Data Papers

The 15th International Conference on Human System Interaction (HSI) is accepting submissions of “Data Papers”. Data Papers and the associated data sets will be fully peer reviewed as it is the intent of HIS 2022 to produce archival quality data.

Data papers are short (circa 4 – 6 pages) papers that support and summarize a substantial archival data set that will be peer reviewed with the same diligence as standard papers.

How to write a Data Paper

Data Papers should be short, around 4 – 6 pages, and will be part of the conference proceedings. Attention should be given to the following:

  • How the data was captured with all relevant detail (e.g., camera calibrations, sensor coordinates, sensor types, timestamp accuracy)
  • How the reader can parse (programmatically) and index into the data
  • Where appropriate, ground truth data should be provided, and details given on how it was obtained
  • Where possible, numerical data should be provided in human readable plain text format, images in high quality jpeg or png format
  • The data should be placed in the context of current research making it clear which research field it applies to
  • If applicable, a commentary on the data is valued. Discuss aspects of the data (or its interpretation) whose processing is within and outside the state of the art at the time of submission
  • The paper should highlight any known issues or obvious difficulties in the data set, e.g., sensor dropout, camera flare
  • Authors are encouraged to provide suitable parsing and access tools in the form of Python/R scripts or any other programming language code
  • Reviewers will be asked to consider the value and quality of the data
  • Papers that seek to describe a new algorithm in conjunction with the data supplied will be considered to be a standard paper, not a Data Paper
  • The paper should be driven from start to end by the desire to make the data used by other authors
  • If the submitted data exceeds 200MB, a representative subset of the data should be submitted for review. Authors should indicate how this sample was chosen
  • If the authors’ submission is intended for an audience/use already addressed by an existing (published) Data Paper, then the authors should be explicit about what shortcomings are addressed by their submission
Review Process

During peer review, the hosting of the datasets should be at the authors chosen site and information of how to access the dataset should be provided alongside the Data Paper. Data papers can be submitted similar to standard papers.

Leave a comment