Data Science Education Meeting

This community site will accumulate resources relevant to developing and implementing programs in data science.  It is an outcome of an NSF-funded workshop, held October 1 - 3, 2015 in Crystal City (Arlington, VA).  The meeting was chaired by Boots Cassel (cassel@acm.org) and Heikki Topi (htopi@bentley.edu) and both are happy to address questions.  

We are now in the process of developing the report of the workshop results.  The workshop brought together a very diverse set of people who approach data science from a number of different perspectives.  The report will attempt to capture the areas of consensus and give due attention to the areas where there are differences of opinion.  The report generation will be an open, collaborative process, with input welcome from all who attended.  Comments from those interested in the topic, but who were not at the meeting, may also be valuable.  

Group: 

Data Science Education Workshop Final Report

This attachment contains version 1 of the report from the October 2015 meeting regarding Data Science Education.  Comments are most welcome.  The report will be presented to the ACM Education Council on August 22.  If you share the report with others, please do it by a link to this site so that we can gather feedback.

Fatal error: Call to undefined function toPrint() in /home/ensemble/ensemble1/modules/php/php.module(80) : eval()'d code on line 3

File Attachments: 
AttachmentSize
PDF icon Data Science Education Workshop Report581.01 KB

2016-7-15: Workshop report as of July 15, 2016 -- Still draft

Draft report from the October 2015 meeting

Initial report outline, open for comment

Initial Outline for Comments Regarding Structure

  • Executive summary
  • Introduction
    • Background
    • Key goals and motivation
  • Charge to the participants and results of the pre-workshop work
  • Structure of the workshop and key topics from each session
  • Thematic findings, not bound by the structure of the event, including the following:
    • Language
      • What is a common language for discussing this?
    • Motivation and momentum of the group
      • This is the time when we have an opportunity to work with something new and exciting.
    • Inherent interdisciplinary nature
      • Not just a combination of courses.  Maybe something new, rather than a combination of disciplines.
    • No clarity regarding whether or not we should define a core
      • Nervousness about defining a core that seemed to box things in.  Generally agreeable that some things are definitely part of data science.
      • Core competences vs. core topics
    • Significant differences in how people see the role of domains in data science
      • Expertise that can be applied to many domains
      • Domain drives the need that data science addresses; data science changes the domain
      • Bioinformatics, business analytics (sometimes data analytics)
        • Evolved in the context of a specific domain
    • Initial thoughts regarding curriculum content
      • See Heikki’s slides that summarize the topics that people identified before the meeting.  Was there any indication of movement of those ideas during the meeting? 
    • Competence requirements vary depending on the stages in the data science process. 
      • Some very abstract thinkers who advance the algorithms and the techniques; others work on data munging, perhaps using tools.  Others do visualization and presentation. 
      • Need or desirability of a deeper understanding of the potential contributions of data science, regardless of the limitations of the eventual role?
    • Changing status quo within universities. 
      • To do it well we need a new structure/environment in the university.  This is the nature of truly interdisciplinary work.
    • Role of ethics and the ability to understand implications for data scientists
    • Integrative conclusion
  • Follow-up actions and recommendations
    • Curriculum
      • Guidance, not constraining
        • Lots of variation supported
        • Some guidance about the elements that any program in data science must include
      • Core competencies?  Expected outcomes?
      • Core topics (Core topics à core competencies)
      • Vote on curriculum recommendation document: all but 2 said yes, 2 said not sure. 
      • Inherently interdisciplinary
        • Not just a combination of courses from a number of disciplines
        • Is it really something new, or a particular combination of existing things?
        • Defining a core may be too much a model of thinking that does not fit this phenomenon.
    • Learning outcomes/objectives for undergraduates who major in data science
    • Building mechanisms to support faculty development 
    • Developing models for collaborative teaching that work with the university scheduling requirements
    • Platform(s) for sharing
      • Exchange of ideas.  Ensemble?  <Umbrella theme to connect multiple sites? Part of our umbrella plan?>
    • Enabling communication regarding things that did not work
      • Establish some sort of venue for publishing stories of lessons learned from efforts that did not achieve their intended goals. 
      • Ethics, implications.
    • Organizing follow-up meetings
      • Involving industry representatives
      • A broader academic/industry group
  • Summary and conclusions
  • Appendices
    • List of attendees and affiliations
    • Agenda
      • List of breakout sessions
      • Summary of results from each group
    • Copies of slides
    • The ASA statement on the role of statistics in data science

Lists of programs

Please add any links that provide lists of data science or analytics programs.  

https://tfetimes.com/2016-business-analytics-rankings/

Report process and timeline

This is our plan for the timeline and process for generating the report for the Data Science meeting. Comments and concerns are welcome.

Please note: The group is open and anyone who is interested can comment on the report as it evolves. If you wish to be able to edit what is shown, you must be a member of the group.  If you are not a membe and would like to be one, send an email to Boots: cassel@acm.org

Report process

10/29/2015 – 11/6/2015: Display a rough outline and ask participants to comment

  • The topics under the headings are representative samples and not intended to be complete.
  • Is something missing?
  • Level of thematic finding, or specific recommendations
  • Comments about the structure.

11/7/2015 – 11/24/2015: Heikki and Boots fill in content, mostly outline form.

11/25/2015 – 12/18/2015: The participants are invited to contribute to the document by 

  • Improving existing content or adding to it
  • Commenting on existing content either by disagreeing with or supporting existing sections
  • Contributing an entire section
  • Commenting on contributions by other participants

1/4/2016 – 1/16/2016: Heikki and Boots produce text for the final report

1/16/2016 – 1/29/2016: Participants are invited to contribute in the following ways

  • Comments and modification recommendations allowed
  • If something seems wrong or not properly represented, we fix and check again. 
  • No new content at that point for the report.

1/30/2016 – 2/29/2016: Boots and Heikki will produce the final version of the report.

The content will remain on the site and the discussion can continue.

DataScience Collections