AIDR Symposium 2020 Submission Instructions

The AIDR 2020 program committee invites abstract submissions for posters and demos that apply AI / machine learning to address challenges related to the discovery, reuse and management of data across disciplinary domains. Innovative algorithms, tools and platforms, interesting use cases, and community building activities are of great interest. This year, we are particularly interested in projects that address challenges in sharing and reusing data related to COVID-19 research.

Submission Deadline: October 11, 2020, 23:59 ET 

Abstracts should be no longer than 500 words. Submissions should be entered in the web form via EasyChair (you will need to have or create an EasyChair account to proceed):

https://easychair.org/conferences/?conf=aidr2020

List of Topics

Topics that the conference will address include but are not limited to:

  • Automating data discovery: Going beyond simple search, how can AI help find data that are described in different ways, languages, and formats?
  • Automating data curation and metadata generation: How can AI provide more robust and more precise tools for provenance, data-driven metadata, (machine-readable) documentation, and curation?
  • Measuring and improving data quality: How can AI help to assess data quality, provide recommendations for improving data quality and develop tools for doing so, and to measure data quality for consideration as it is being reused?
  • Integrating datasets: As relevant data are discovered, how can AI help with their fusion, including factors such as ontology, format, units, and language?
  • Enabling interpretability: How can AI contribute to the representation of information that is both machine readable and human readable?
  • Measuring data metrics and citation: How are data citations tracked and linked to publications? How is the impact of research data evaluated?
  • Data privacy, security and algorithmic bias: what are their ethical implications and how can they be avoided?
  • Collaborating across disciplinary and expertise domains: How do professionals from different domains work together?

Areas with great potential to address these challenges include but are not limited to:

  • Natural language processing to aid in the discovery of data and its interpretation, information extraction, and generation of ontology, taxonomy and the knowledge base
  • Inference of data types and where they fit into ontologies, and automatically creating more precise, machine-readable metadata from that information
  • Measuring, reporting, and improving data quality through identification and potential cleaning of missing or possibly erroneous values
  • Inference-based conversion of formats, ranging from simple cases such as unit conversions to more complex cases such as working with data represented using different geographical frames
  • Tools for visually representing data quickly and intuitively, to help with understanding unfamiliar datasets and the results of analytics
  • Human-in-the-loop methods applied to all of the above for semi-supervised training, potentially leading to greater degrees of autonomy

Contact

All questions regarding submissions should be emailed to aidr@andrew.cmu.edu

For more information and to register, visit the event website: https://events.library.cmu.edu/aidr2020/