The document discusses the need for archaeological information systems to model concepts and ideas from archaeological data in a practical way. It advocates building on existing standards like CIDOC-CRM and CRM-EH to create robust data models that can represent temporality, uncertainty, subjectivity, and multivocality in archaeological data. Recent experiences at Wessex Archaeology developing spatial data infrastructure approaches based on events, people, places, and stuff are provided as an example.
Places, People, Events and Stuff; building blocks for archaeological information systems
1. Wessex Archaeology
Places, People, Events and
Stuff; building blocks for
archaeological information
systems
Concepts and ideas to practical
implementation
Paul Cripps
•Geomatics Manager, Wessex Archaeology
•Archaeological Computing Research Group, University of Southampton
http://www.wessexarch.co.uk/geomatics Computer Applications & Quantitative Methods in Archaeology. Southampton. March 2012.
2. Wessex Archaeology
Overview
• Why?
– The point of archaeological
information modelling
• How?
– Building on what is out
there
– Low hanging fruit and all
that…
– Getting stuck in
• Recent experiences @
Wessex Archaeology
Image courtesy of Sophia Yip: sophiayip.com
http://www.wessexarch.co.uk/geomatics Computer Applications & Quantitative Methods in Archaeology. Southampton. March 2012.
3. Wessex Archaeology
Why do we need to model…?
• Basic database standards/principles for effective
data management
– Redundancy
– Duplication
– Normalisation/Denormalisation
• Semantic clarity for search/retrieval/analysis
– What does ‘Roman’ actually mean…?
• Interoperability
– Is my ‘Roman’ the same as your ‘Roman’
http://www.wessexarch.co.uk/geomatics Computer Applications & Quantitative Methods in Archaeology. Southampton. March 2012.
4. Wessex Archaeology
Why do we need to model…?
• Problems:
• Too often, a perfect view of the world
– Basic models do not support richness of data
– Do not support change management ie iterative
assessment/analysis
• Semantically unclear
– eg ‘Period’ ascriptions
• Also fundamental database issues
– Lack of atomicity
– Generation and use of IDs and keys eg SMR numbers
• Unnecessary/undefined/undocumented complexity
– Can lead to inconsistency in use as users are unclear about how
to proceed
http://www.wessexarch.co.uk/geomatics Computer Applications & Quantitative Methods in Archaeology. Southampton. March 2012.
5. Wessex Archaeology
Why do we need to model…?
• Aims:
• Information Systems that appropriately represent the
archaeological record & support its maintenance
– Variable in quality
– Provenance
– Presence/Absence; knowing the unknowns
• Support & enhance the archaeological process
– Inference
– Evidence
– Multi-vocality
• Get away from rudimentary, poorly structured systems
– A hindrance not a help
http://www.wessexarch.co.uk/geomatics Computer Applications & Quantitative Methods in Archaeology. Southampton. March 2012.
6. Wessex Archaeology
How do/can we model…?
• Standard database driven techniques
– Conceptual, Logical, Physical
• Entity-Relationship models
– Great but can be limited
– Easy to implement
• Object-Oriented models
– Powerful but can be complex
– Harder to implement
• Object-Relational models
– Best of both…?
– OO concepts in a relational DBMS
http://www.wessexarch.co.uk/geomatics Computer Applications & Quantitative Methods in Archaeology. Southampton. March 2012.
7. Wessex Archaeology
How do/can we model…?
• Event driven models • Events in the Present
– Any data object is the product – ie archaeologists (people)
of an Event doing archaeology
– Very useful for describing the • Events in the Past
archaeological process – ie past peoples living,
• Can reduce everything to a experiencing, interacting with
few core elements each other and the world
• Typologies, classifications around them; leaving…
– A large proportion of what we • Stuff
do – ie archaeological remains;
• Object Inheritance finds, structures, etc in…
– Subtyping • Places
– Ensures robust data objects – ie depositional contexts,
structures, geographic
entities, etc
http://www.wessexarch.co.uk/geomatics Computer Applications & Quantitative Methods in Archaeology. Southampton. March 2012.
8. Wessex Archaeology
Is all this really necessary…?
• Why not just publish Linked(Open)Data from
existing information systems…?
– Put it all out there, use will follow
– It’ll all come out in the wash
– Mashups
• It’s really complicated, lot’s of work involved…
– Serious investment for limited gain
– Why not focus on the low hanging fruit…?
http://www.wessexarch.co.uk/geomatics Computer Applications & Quantitative Methods in Archaeology. Southampton. March 2012.
9. Wessex Archaeology
Yes, it is really necessary!
• The basics: effective data management
• Linked(Open)Data works better with good data
– Semantic inconsistencies problematic
– Great for delivering and sharing data but not a solution in and
of itself
• Focussing on the low hanging fruit misses bigger
potential
– Yes, by all means go for it!
– But, complex data (eg archaeological excavation data) requires
more complex solutions
– But, we still need to work towards semantically clear, truly
interoperable information systems
– Potential for use of techniques such as Natural Langauge
Processing, Crowd Sourcing to populate suitable data models
http://www.wessexarch.co.uk/geomatics Computer Applications & Quantitative Methods in Archaeology. Southampton. March 2012.
10. Wessex Archaeology
Moving forwards
• Obviously proceed with easy targets
– Leverage eg Google, LinkedData initiatives, etc
• But also build on work to date:
• CIDOC-CRM
– Not aimed at systems design but useful concepts therein
– ISO standard
• CRM-EH extensions
– Extensions to the CIDOC-CRM for archaeological (fieldwork) data
– Models big chunk of the archaeological process
• Star + Stellar projects
– Tools for working with CRM-EH
– Broad range of contributors
• Ongoing & forthcoming projects
– eg KOS representations of thesauri to populate information
systems
http://www.wessexarch.co.uk/geomatics Computer Applications & Quantitative Methods in Archaeology. Southampton. March 2012.
11. Wessex Archaeology
Some specifics
• Concepts
– Subjectivity
– Multivocality
– Temporality
– Uncertainty
• Ideas for tackling these areas
http://www.wessexarch.co.uk/geomatics Computer Applications & Quantitative Methods in Archaeology. Southampton. March 2012.
12. Wessex Archaeology
Subjectivity & Multivocality
• All assertions are the product of an Event
– Phasing & Dating
– Classification
• All assertions are made by a Person
– Multiple archaeologists = multiple stories
• All assertions based on evidence
– Stuff originating from Places
• Review & Confidence ascriptions as Events
– Explicit inference within the information model
http://www.wessexarch.co.uk/geomatics Computer Applications & Quantitative Methods in Archaeology. Southampton. March 2012.
13. Wessex Archaeology
Temporality & Uncertainty
• Events in the past have • is equal in time to
associated time-spans • finishes (is finished by)
• Temporal reasoning using • starts (is started by)
Allen Operators • occurs during (includes)
• Uncertainty about time • overlaps in time with (is
can be modelled using overlapped in time by)
these operators
• meets in time with (is
– Phasing & stratigraphy;
relative chronology met in time by)
– Scientific Dating; absolute • occurs before (occurs
chronology after)
http://www.wessexarch.co.uk/geomatics Computer Applications & Quantitative Methods in Archaeology. Southampton. March 2012.
14. Wessex Archaeology
Experiences from the coal face
• Wessex Archaeology currently developing SDI approaches to digital
(spatial) data
• Based around robust data/process model: Events, People, Places,
Stuff
• Multivocality, assertion & inference, uncertainty
• Publishing/archiving/disseminating as CRM-EH RDF, LinkedData, WMS/
WFS, etc
– One project ongoing
– Aim: all WA data currently in digital form (= many sites, some of which
are massive!)
• Potential for external linkages to other fieldwork datasets for analysis
– IADB
– Intrasis
• Potential for linkage to museum collections
– MODES
• Potential for linkage to HER/SMR webservices
– HBSMR
http://www.wessexarch.co.uk/geomatics Computer Applications & Quantitative Methods in Archaeology. Southampton. March 2012.
15. Wessex Archaeology
Thanks!
• For more information please contact me:
– p.cripps@wessexarch.co.uk
– wessexarch.co.uk/geomatics
http://www.wessexarch.co.uk/geomatics Computer Applications & Quantitative Methods in Archaeology. Southampton. March 2012.