Spatial Data Science

Keynote Response
Dani Arribas-Bel [@darribas]

The Rise of Data Science

Data, Data, Data

Every-thing produces data

New, dense, quantitative representations of the world

But data is not useful, insights are

Data Science

Data Science

gathering data, massaging it into a tractable form, making it tell its story, and presenting that story to others.”

Loukides (2011)

Hold on…

  • A lot of the new data is spatial data
  • Spatial is special
  • We don’t want to reinvent the (GISc) wheel
How can we bring the best of both worlds together?

Geo/Spatial Data Science

Why?

Fully tap into the new data revolution

  • Data are different (“half-baked”)
  • New data allow more flexibility (+ data, - structure)

“Business-as-usual” best approach?

A (rough) sketch

What

  • Systems engineering [Geo-infrastructure]
  • Methods [Explicitly spatial ML/Stats]
  • Epistemology [Adequate conceptualization]

How

  • Coupling [Linking Platforms]
  • Assimilation [space fully embedded in DS]
  • Full Integration [Co-Proudction]

Potential Benefits

  • Fully leverage the new data revolution
  • Do not reinvent the wheel
  • New answers to old questions (e.g. MAUP)
  • Old answers in new contexts
  • New questions to answer (new methods?)

Getting everyone behind G/SDS

G/SDS is for all

Industry & Academia

  • Shared problems
    • Access to (/baking of) data
    • Similar questions in different contexts
  • Complementary resources
    • [A] Training muscle, space to experiment
    • [I] Funding muscle, production-ready tooling

Clear synergies

Challenges

  • Communication what’s needed/valued?
  • Interaction how?
  • Credit Fame-sharing

For example…

  • Teach relevant G/SDS skills
  • Provide close interaction w/ industry in training
  • Build platforms that enabling data sharing/access

For example…

  • Teach relevant G/SDS skills
  • Provide close interaction w/ industry in training
  • Build platforms that enabling data sharing/access

For example…

  • Teach relevant G/SDS skills
  • Provide close interaction w/ industry in training
  • Build platforms that enabling data sharing/access

For example…

  • Teach relevant G/SDS skills
  • Provide close interaction w/ industry in training
  • Build platforms that enabling data sharing/access

Spatial Data Science

Keynote Response
Dani Arribas-Bel [@darribas]