This document discusses the need for new metrics to measure the impact and value of open data beyond traditional metrics like usage, downloads, and visualizations. It proposes a new model for data metrics that examines data across three dimensions: spatial (location), temporal (time), and topical (subject matter). For each dimension, examples are provided like analyzing how data relates to neighboring datasets, its history over time, or topics through ontologies. The model is applied to Pablo Picasso's painting Guernica to demonstrate how its spatial, temporal, and topical attributes could be measured. Feedback is sought on whether these three dimensions capture enough context or are too complex, and how well different open data sources incorporate these dimensions.
10. A new model for Data Metrics
The dimensions on which we can work on this model can be
š Spatial (do neighbours have a dataset?)
š Topic-driven (do we talk about the same stuff?)
š Time-bound (do we have a history about that data?)
11. A new model for Data Metrics - Space
Cool! Very Uncool L
12. A new model for Data Metrics - Time
Time series (EuroStat, ISTAT, National Statistics agencies) J
Anyone else? L Municipalities have lots of historical data!
13. A new model for Data Metrics - Topic
š Ontologies! à We have lots of people working on them!!
š We have ENORMOUS amounts of CSVs…
š Can we give semantics to everything? SURE!!
15. Let’s see Guernica…
š Spatial information:
š Guernica is about Guernica (Location)
š Guernica is in Madrid (Location)
š Guernica was painted in Paris (Location)
š Time information:
š Guernica was painted in 1937
š Guernica was exposed at the MoMA until September 1981
š Guernica was exposed at the Casón del Buen Retiro until 1992
š Guernica is exposed at the Museo Reina Sofía
š Topic Information:
š Guernica is about War, pain,
š Guernica is about Guernica (Location)
š Guernica is about The bombing of Guernica (Event in space and time)
š Guernica is enormously complex: http://en.wikipedia.org/wiki/Guernica_(painting)#The_painting
16. So now what?
š We now have the three dimensions. Are they enough? Are they too much?
š Europeana: has all of them (in part) [depending on the availability of the single institutions]
š City of Chicago: Has all of them and they are pretty well structured [lots of id columns enabling
foreing keys]
š Most datasets: L