ann zimmerman 1075 beal ave. ann arbor, mi 48109-2112 asz@@umich.edu the local goes global how do...

Ann Zimmerman1075 Beal Ave.Ann Arbor, MI 48109-2112asz@@umich.edu

The Local Goes GlobalHow do data come to be shared?

Ann ZimmermanResearch Assistant Professor

ASIST Data Summit, April 10, 2010

Ann Zimmerman1075 Beal AvenueAnn Arbor, MI 48109-2112asz@umich.edu

Data collections

•Research

•Community

•Reference

Ann Zimmerman1075 Beal AvenueAnn Arbor, MI 48109-2112asz@umich.edu

Local project

Share Publish Create Reference Collection

Ann Zimmerman1075 Beal AvenueAnn Arbor, MI 48109-2112asz@umich.edu

• Why is it hard to share data?

• Why is data sharing more common in some fields than others?

Science-based

Technical

Organizational

Social

Legal

Political

Ann Zimmerman1075 Beal AvenueAnn Arbor, MI 48109-2112asz@umich.edu

THE LOCAL GOES GLOBAL: SOME EXAMPLES

Ann Zimmerman1075 Beal AvenueAnn Arbor, MI 48109-2112asz@umich.edu

Implications

• Who is involved

• What count as data

• What gets shared

Ann Zimmerman1075 Beal AvenueAnn Arbor, MI 48109-2112asz@umich.edu

Step 1: Scientist publishes a a paper in a journal.

Step 2: A data curator scans journals for data, extracts data and descriptive information, and enters

it all into a computer.

Step 3: Data are integrated with other data into one database and made available to anyone via the

Internet.

Step 4: Other people use the data. They provide input that results in additions of new types of data or

corrections to the database.

Step 1: Scientist publishes a a paper in a journal.

Step 2: Scientist submits the data associated with the paper to a repository as a requirement of

publication.

Step 3: Data are integrated with other data into one database and made available to anyone via the

Internet.

Step 4: Other people use the data. They provide input that results in additions of new types of data or

corrections to the database.

Step 1: Individuals get together to decide upon standard data

collection protocols.

Step 2: Every laboratory uses the agreed upon methods to collect data.

Step 3: Data are integrated and made available to anyone via the

Internet.

Step 4: Other people use the data. They provide input that influences how new data are collected or that

result in corrections to existing data.

Ann Zimmerman1075 Beal AvenueAnn Arbor, MI 48109-2112asz@umich.edu

Implications

• Who is involved

• What count as data

• What gets shared

Ann Zimmerman1075 Beal AvenueAnn Arbor, MI 48109-2112asz@umich.edu

THE LOCAL STRUGGLES TO GO GLOBAL

Ann Zimmerman1075 Beal AvenueAnn Arbor, MI 48109-2112asz@umich.edu

Materials science data

Images

Graphs

Spectra

Columns of numbers

Ann Zimmerman1075 Beal AvenueAnn Arbor, MI 48109-2112asz@umich.edu

Shared Needs

• Long term access to data

• Finding data later

• Understanding the context of the data to use it in a meaningful way

• Accessing data, and information about data, from multiple locations

Ann Zimmerman1075 Beal AvenueAnn Arbor, MI 48109-2112asz@umich.edu

Open Questions

• What makes it hard to share data?

• How does the “state” of data affect sharing?

• What makes documentation sufficient for reuse?

Ann Zimmerman1075 Beal AvenueAnn Arbor, MI 48109-2112asz@umich.edu

Developing Capacity

• Open Data IGERT (Margaret Hedstrom) http://opendata.si.umich.edu

• “i-School” Masters programs

Ann Zimmerman1075 Beal AvenueAnn Arbor, MI 48109-2112asz@umich.edu

Acknowledgments• Data Summit organizers &

attendees

• Research participants

• Dharma Akmon & Morgan Daniels, PhD students, UM School of Information

• NSF Grants OCI 0724300 and IIS 0085981 (Gary Olson, PI)