The role and value of making data inventories a key step towards - - PowerPoint PPT Presentation

The role and value of making data inventories

a key step towards mature data governance #openbelgium Louvain-la-Neuve,12 March 2018 Ton Zijlstra, @ton_zylstra, thegreenland.eu, slides: https://grnl.eu/in

The role and value of making data inventories

Province Utrecht Province Fryslân Province North-Holland City Eindhoven City Leeuwarden City Delft

1 high time for mature data governance

www.flickr.com/photos/31954002@N08/14811288593/

digital isn’t paper redone

digital changes how we look at

• openness: access is different from re-use
• privacy: different types of usage
• security: non-binary
• archiving: earlier in information processes

three fences

security

three fences

security

• penness

three fences

security

• penness

privacy

three fences

security fence

• ‘baseline information security’ for local/regional govs

determines what data is critical

• Uptime: IT infrastructure, dependencies, service levels
• Quality: tamper free, audit processes, checks on inputs,

knowing sources

• Uptime fits the fence tactic, quality doesn’t

• penness fence
• at request, additional process
• stated end game is ‘actively open by design’
• open data is becoming infrastructure (e.g. ‘omgevingswet’)
• fence tactic is inefficient

• data sovereignty is under threat
• not enough attention at data level
• fence tactic is ineffective
• penness fence

• complaints about

compliance costs

• house not in order
• PSI Directive Review

confirms

• penness fence

www.flickr.com/photos/wfabry/2157854271/

• done on level of organisation or system
• GDPR is here, creates uncertainty
• excuse for ‘closed by design’
• making lists for the fence, not processes

privacy fence

• right to review
• right to portability
• right to be forgotten (archiving)
• “by design, and state of the art”, and is enforced
• can only be done at data level, and processes tapping

into data

• the fence tactic fails completely

• GDPR demands ‘by design’
• no sense on its own
• openness, archiving, security (Q) as well
• focus on data

GDPR opportunity for ‘everything by design’

• pen

30 yr limit person related 3rd party rights business critical

data focus & ‘everything by design’, not fences

2 local data inventories a first step

local is where you are, but not the data pro’s

value, impact

policy issue people

• pen data

connecting people and issues needs data knowledge

Actief uitnodigend

published inventory triggers demand

http://www.flickr.com/photos/dhammza/492882480/

new local data means new relations & choices

inventories help having the right conversations

external stakeholders data person policy maker 3rd party internal stakeholders domain specialist legal person

https://www.flickr.com/photos/mikecogh/11300349426

3 process and results

starting points

• list structured data sets only
• up to 80 facets
• policy domain, internal usage, current availability,

technical details, legal aspects, and concerns

don’t make assumptions, because house not in order

• don’t assume your list of applications will tell you
• don’t assume IM knows
• don’t assume people know
• don’t assume people know details, look inside with them
• don’t assume it is what it says on the tin

• actually used applications
• all units, actual work
• data structures and content
• large projects / programs
• external communications
• cross reference it all

deep dive

www.flickr.com/photos/jphotos/5945632837/

tactic 1: external team

• consistent, no

assumptions

• experience
• re-use of results
• client buy-in can be low
• hand-over can be hard
• the work is not an

intervention itself

tactic 2: ext./client team

• train client team
• consistency
• experience & re-use
• needs more scripted

approach

• quality can be issue
• client team continuity
• islands likely remain

tactic 3: client team / employees

• train client team
• very scripted approach
• process facilitators
• adoption designed into

process

• work shifted to

colleagues

• general buy-in critical
• quality output trade-off
• no guaranteed adoption

28

Typical situation (local 67%, <5%, 33%)

Data inventory Province Fryslân 2016

• 1055 data sets found (767 geo)
• 201 public (19%), of which 151 (14%) open

data (all geo)

• 841 more could be public (79%), after

changes (304, 29%)

• 17 (2%) must stay closed

allows filtering on all relevant questions

published, and used to select next publication round

www.flickr.com/photos/dteslya/4254871326/

Legal (GDPR, infosec) IT/architecture (infosec) Data people (data q, openness) Archiving Policy people (openness as instrument)

involving all from start helps handover

summary

• information household is often of poor quality
• tear down the ‘fences’
• inventories help make a start, if you see it as a

conversation tool not just another list

• helps connect ‘everything’ by design, as step towards

mature data governance

• articulates demand, allows data as policy instrument

Thank you. Merci. Hartelijk dank.

All photos: Ton Zijlstra, by
 Except screenshots, and where mentioned on the photo. Slides: Ton Zijlstra / The Green Land, by nc sa Slides: https://grnl.eu/in Site: https://thegreenland.eu Contact: ton@thegreenland.eu @ton_zylstra