[Colloquium] Seminar Announcement: Adaptable and Incremental Metadata Capture in e-Science

[Ninfa Mayorga]

~Reminder~

Computation Institute Presentation - Data Lunch Seminar (DLS)

Speaker: Scott Jensen, Postdoctoral Researcher, Data to Insight, Indiana University
Host: Tanu Malik 
Date: March 2, 2012
Time: 12:00 PM - 1:00 PM
Location: University of Chicago, Searle 240A, 5735 S. Ellis Avenue

Adaptable and Incremental Metadata Capture in e-Science

Abstract:
Scientific communities are recognizing an increasing need to enable reuse of the deluge (or bonanza) of scientific data currently being generated. Detailed metadata, or 'data about data', is key to preserving the value, as well as enabling the sharing and reuse of data. Communities have developed detailed XML schemata to capture and communicate metadata describing scientific data. Historically however, to the extent metadata has been captured at all, it was done at the end of an experiment when results are published and being curated. This approach does not scale well with the increasing volume of data being generated and results in much of the metadata needed to understand a data object being lost due to the ephemeral nature of metadata itself. 

To address these issues, we push metadata capture to the earliest stages of the scientific data lifecycle as data objects are created and the scientist generating the data is the steward of the data. However, scientists often see little benefit to documenting their data with metadata, in part due to a misalignment of incentives between those generating metadata and future users benefiting from the reuse enabled by metadata. In this talk I will discuss how we are exploiting characteristics of scientific metadata schemata to enable the efficient incremental and automated capture of detailed metadata. This approach reduces the misalignment of incentives by reducing the burden on the scientist through automation while also increasing the utility of the metadata to the original researcher by making it available during an experiment's execution. This approach uses a generalized underlying architecture that can be applied across the schemata of different scientific communities and 'talk' the metadata schema of the community implementing the system.

Bio:
Scott Jensen is currently a post-doctoral researcher in the Data to Insight Center at Indiana University. He received his PhD. from Indiana University, Bloomington in 2010. His research focus is on capturing the metadata and provenance needed to enable the reuse of scientific data and the leveraging of data across scientific disciplines. His research interests also include web services, XML, XML-relational data storage, data search, and the semantic web. Scott earned an M.S. in Computer Science at DePaul University in 1996, and a B.S. and MAcc. from Southern Illinois University in 1984 and 1986 respectively.


Information: Lunch will be provided