Discussion Notes

Open Data Policy and Licensing

Open Data for Business

Potential drivers

• Everyone benefits from “big data”
• Cost saving for non-competitive data
• Competitive advantage
• Could possibly create a business model providing high quality data
• Emerging companies

New applications / business models

• Pharma research can benefit from academia collaboration
• Broker applications (insurance, travel etc…)
• Event data: link to transport data; weather
• Meter data: link to GPS
• Any website with a shopping cart (product/service, etc)
• Journals that made data publicly available have benefited the most
• Health and safety data within the company
• E.g. tripadvisor gives companies the chance to go back and apologize/fix negative reviews
• Ability to integrate data across rating sites
• Standards organizations (ISO, etc)
• Weather and road information (Visualisation of floods)
• Open your data to ensure that your company is featured in Search Engines
• Airline business model
• EMI Business model
• Could learn from UK open data corporation

Why this isn’t happening

• Fear of the unknown
• Afraid of being compared
• Business are focusing more on business process efficiency
• For the private sector open data is more of an interest
• Maybe not suitable for everyone
• Can’t afford to do it
• Don’t realise not sharing data will ruin their business
• IP issues
• Sharing data imposes quality control
• There is a need for standards

Open Data Quality and Standards

Participants

• Trevor (Marine Institute)
• Liam (Marine Institute)
• Peadar (Dublin Institute of Technology)
• Aftab (DERI)
• Freddy (IBM)
• Peter (UCC)
• Atif (Trinity)
• Eoin (CSO)
• Sarven (DERI)
• Lin (DERI)
• Michael (DERI)

Discussion

• Quality - when is the right time?
• Is quality an objective measure? Depending on the use case - hard facts vs. soft facts (metadata available?)
• One measure is how effectively the reality is modelled
• Taking data management issues into account ('auditable' ?)
• Correct data vs. useful data
• Feedback loop is essential
• Marine environment there are a number of quality assurance
  • through external experts)
  • information loss through throwing away data that is beyond expected range
  • need for correction (due to faulty measurements, etc.)
• time dependencies
  • transactional reality vs. physical reality)
  • TIE example / legal resources
  • capture in the model vs. formats
  • numerical data vs. textual data
• Dublinked - need for data curation/cleansing (ODO responsible for quality)
• INSPIRE does not provide direct quality conformance, it's about metadata
• who's responsibility is quality? consumer? ODO?
  • INSPIRE - legal mandated organisation - provenance is essential to trace back
  • need for different roles in the organisation (chain of authority across technical and content fields)
  • review of data production in place for statistics
  • adopting ISO9001 QM for data publishing (?)
• different types of data: collected for a special purpose and generated
• estimates vs. accurately captured
• Idea: to motivate data publisher - 'data amnesty' - dependent on requirements (?)