Documentation - February 2014 Telecon

1. Strategic plan (https://docs.google.com/document/d/10lKi88rReL0rF4WhZ8yEMabBoQTuALX-vu2VXBVokCo/edit?usp=sharing)

·         Looks finalized

·         (Ed) should we move forward with the hack-a-thon

·         ACTION – Anna will submit a session for the summer meeting for the hack-a-thon

·   Need a conceptual framework for the hack-a-thon (structure)

·   ½ day (3 hr)?

·   People would bring their own hardware & metadata

·   Tie it into a particular spiral or goal for metadata

·   ISO 19157 – data quality group in ESIP… new and has interesting capabilities for data quality

·         Provide an example to work through – have people volunteer their data sets and then in teams work to document the quality and then reviewing together

·         Erin – this sounds more like “training”

·         Anna – might have more interest if more open than training

·         Ted – 3 or 4 data quality use cases (such as Ed’s example of a satellite granule, data and quality information for swath)

·         In 19157 includes a stand-alone quality report

·         Introduction to iso quality metadata and then bring your data quality examples for the second session (2 sessions)

·   GRIST granules has lots of good examples (Ed)

·   Building off resources for the community and their data sets

·         ACTION – Kelly will put strategic plan on the Commons

·   Question about the ESIP values in the strategic plan

·   Erin – context of how fit into ESIP at large

2. "Drop It! Good Boy! (and Other Metadata Strategies)" by John Graybeal 

·         This is about motivations and motivating good metadata into systems

·   People’s motivations can be different – need to take steps to help people reach their goals to meet our goals

·         Start with reality of cost and benefits

·   Cost  - Time, money, privacy/secrecy, simple life, free will, independence

·   Benefits – publicity, functionality, reusability, interoperability, understanding, social grace

·   When you first enter metadata – costs are immediate (time, money)

·   The benefits are delayed.  Maybe immediate publicity

·   Costs are very concrete (measureable).  The benefits are less concrete (they are abstractions)

·   Balance is lop sided because of the costs

·         Change ahead – more integrated systems are coming online – trying to change the cost-balance equations

·         Mandate – ex like commanding your dog to drop something

·   When mandate is strong and well enough enforced they people do it (like taxes)

·   2 examples

·   fgdc – they were bringing in csdgm and trying to get it adopted by all providers of environmental data sets – had mixed uptake

·   inspire – has had astonishing amount of uptake of metadata requirements

·   not just about providing a mandate

·         Other motivations – social persuasions

·   on marine explore (marinexplore.org) – uses points and achievements

·   tangible rewards – ex. Being able to use the system

·   ex. LinkedIn – need to provide information to get at more features

·   in marineos (the private installation of marine explore) – you can see what you have provided

·         Recommendations

·   Why do you care about making the metadata accessible

·   People will provide information because they have to or because it provides value to them

·   How can I had more value into my system

·   Need…. more data, more data functions, more data relations

·   Thus need more better metadata

·   Minimize work

·   Autoimport (& API), Only ask once, Autosuggest/complete (whenever possible… control vocabulary), Autoverify, Minimize keys/clicks, Visual guidance (make things clean and easy to follow), help and examples, prioritize fields (not always able to fill all of it), normalize/autolookup, templatize/autolookup

·   Maximize reward

·   Show progress (feedback at every step, increase social rewards, overall competition), show reuse of metadata, show benefits on the fly (relations/groups of data, possible overlaps/duplicates, similar data sets/providers)

·   Make the process fun!

·         Ed – any comment on prioritizing the mimize work and maximize rewards column

·   1) auto import 2) only fill in once

·         Anna – liked reward – help you understand your data better

·   Find the researchers (for example) don’t completely understand the units that their data was collected in

·         Ted – talk about metadata in relation to “trust”

·   Can talk about scales of trust

 

3. Metadata of the Month – Ted – Identifying Metadata records

·         Significant change – identifier for the metadata record – it was a character string (unstructured)

·   At NGDC put namespaces into character strings

·   It didn’t use MD_Identifier class

·   NOW – it is an Identifier … it is a code, an authority for the code, a code space, and a description

·   More robust

·         Concept of parent and child … in old it was a character string and is now a citation

·   This can include a url for the metadata

·         Stephen – usually use GUI (globally unique identifier) or UUIDs – but they identify the record themselves and some use the same for metadata and data…

·   John – if it is a NetCDF file with the data in the same file then it works

·   Ted – think there are 2 pieces – 1 technical discussion of UUI, DOI, …  and then the social benefits

 

Rich Signell - https://github.com/OSGeo/Cat-Interop/blob/master/ServiceTypes.md