ADMIRAL 20100219 meeting Data Surveys Review

From ImageWeb

Data Surveys reviews

Meeting to discuss initial data usage survey returns from Silk and Development groups.

Date: 19-Feb-2010

Present: DS / DG/ GK

Goals:

• Prioritize initial services to provide for research users
• Identify topics for 1:1 discussions with researchers
• Identify any improvements to the survey

Discussion

Survey returns indicate our original priorities are about right.

There was some discussion of the role of metadata/annotation capture. The surveys focused attention of subsequent retrieval of data, but the overall project aims also include capture for repository submission and interpretation by other researchers. We agreed that the current focus on retrieval is seen as useful by researchers, and that this is appropriate as a near-term focus. This has mimplications for how we may choose to handle the metadata (e.g. provision of indexing and sercah facilities).

We had not really come to understand the required data access patterns; e.g. how much data would be accessed and/or updated frequently vs how much would be occasionally accessed.

Outcomes

Current development priorities continue to be:

• provision of group-level shared filestores
• access control to shared filestore
• automated backup of shared filestore to HFS
• Metadata and annotation capture for keyword-based retrieval

Topics for immediate discussion with researchers:

• data volumes and access patterns; e.g. how much data is (a) frequently updated, (b) frequently accessed, (c) required to be available for infrequent online access, (d) required to be preserved for possibly-delayed access (e.g. from HFS)?
• how much new data do they expect heir group to create over a three-year period, or how much per year (e.g. how to interpret AK's 0.5Tb/project)
• how do researchers' feel their use of the shared filestore interact with their personal data collection and use; e.g. all data saved directly to and accessed from shared filestore, all data saved and accessed locally, and MANUALLY copied to the shared filestore when appropriate, or data saved and accessed locally, and automatically copied to the shared filestore?
• which version of Internet Explorer is used?
• do they experience difficulties setting up automatic backup to HFS?

Possible changes to the survey include:

• questions about ease of accessing/understanding research data should include colleagues as well as group leaders
• Ask for system and browser versions used (especially IE)
• (others? I think some small tweaks, but don't recall details)