Transcript A Science Fiction Tale

This is NOT a Science Fiction Tale
Expert Systems in Archaeology and Culture Heritage.
20 years later
Is it possible to design a
machine to do archaeology?
Will this machine be capable
of acting like a scientist?
Was it used as a bottle?
What’s this?
Was it used as an arrow point?
YES
THINK(rationally), PERCEIVE-EXPLAIN
Expert System
• A computer system that
is programed to mimic the
procedures and decisions
that "experts" make.
• A domain specific
knowledge base
combined with an
inference engine that
processes knowledge
encoded in the knowledge
base to respond to a
user's request for advice
12
The Structure of Archaeological
Intelligence
EXPLANATIONS
DATA
Jean Claude Gardin:
“Une Archéologie Théorique”
(1979)
Rules
•
•
If
(x,y,z) are proper empirical features of Object F1
And (v,w) are proper definition terms of Concept F
Or there is some contextual similarity between F and F1
Then F1 activates F
Object (F1) is an instance of Concept (F)
EXAMPLES:
IF
(x) is a settlement
And
(x) has (y) in quantity (h)
And
(y) is an object of ceramics
or
(y) is a glassware
And
(y) is dated in the 10 th century BC
THEN VERIFY THE ORIGIN OF (y)
IF
(Goal) is TO VERIFY THE ORIGIN OF (y)
And
(y) is made of foreign material
THEN (y) is an Imported Object
IF
(y) is an Imported Object
And
(y) is similar to the Muslim pottery from the Castle of Silves (Portugal)
THEN (x) has Foreign Trade evidence.
EXPERT SYSTEMS IN
ARCHAEOLOGY
A REVIEW OF CURRENT APPLICATIONS
LITHAN (LITHic ANalysis of stone tools)
http://www.hf.uio.no/iakk/roger/lithic/expsys.html.
if
If length/width ratio >2
and width <12 mm.
then put "BLADE LET"
and
and
and
then
if
diff (length - width) > 0
and distalRetouch = "DISTAL"
then put "END SCRAPER"
If
and
and
then
if
then
endForm = "ROUND"
put "END SCRAPER"
platform Thickness <5
ButtType = "prepared"
Sides = "parallel"
Ridges = "parallel"
put "TECHBLADE"
percussionCone = "no cone"
butt = "un-lipped"
bulb = "diffuse"
put "SOFT HAMMER"
if
then
endForm = "CARINATED"
put "CARINATED END SCRAPER"
AVIAN ARCHAEOZOOLOGY
Archaeological Applications of
Expert Systems Technology
•
•
•
•
•
•
•
•
•
•
•
LITHIC ANALYSIS AND DETERMINATION
ARCHAEOZOOLOGY: ANIMAL BONES DETERMINATION
TYPOLOGY
ANCIENT WOOD TAXONOMY
ARCHAEOMETRY: PROVENANCE STUDIES
RECONSTRUCTION
EXPLAINING DECORATIVE PATTERNS
EPIGRAPHY AND ANCIENT TEXTS
MUSEOLOGICAL AND CURATOR STUDIES
GEOSCIENTIFIC PROBLEMS
SOCIAL AND HISTORICAL INTERPRETATION
data ≠ knowledge
The impossibility of “artificial” intelligence?
I’m sorry. I have no answer. Nobody
taught me I should learn that solution
THE TASK:
to find the common
structure in a given
perceptual sequence
ASSUMPTION:
the structure that is common
across many individual instances
of the same cause-effect
relationship must be definitive of
that group
OBSERVATION OF INDIVIDUAL INSTANCES
Cause
Cause
Effect
Effect
...
...
Cause
Cause
Effect
Effect
C
O
M
M
U
N
A
L
I
T
I
E
S
+
PRIOR KNOWLEDGE:
constraints that will ensure that the predictions drawn
by an automated archaeologist will tend to be plausible and relevant to the
system’s goals
testing
Feedback cycle
Description
Explanation
Inference of a general model
Simulation
Bayesian networks
NEURAL NETWORKS
SOFT CLASSIFICATION
FUZZY LOGIC MODULE
LOHSE, E.S., SCHOU,C., SCHLADER,R., SAMMONS,D, 2004, “Automated Classification of Stone
projectile Points in a Neural Network”. In Enter the Past. The e-way into the four dimensions of culture
heritage. Edited by Magistrat der Stadt Wien-Referat Kulturelles Erbe-Städtarhchäologie Wien. Oxford,
ArcheoPress (B AR Int. Series, S1227), pp. 431-437).
J.A. BARCELO, 1995 Back-propagation algorithms to compute similarity
relationships among archaeological artifacts. In Computer Applications in Archaeology.
Edited By r J. Wilcock y K. Lockyear. Oxford: British Archaeological Reports.
DIAZ,D., CASTRO,D., 2001, “Pattern Recognition applied to Rock Art”.
In Archaeological Informatics: Pushing the Envelope. Edited by Göran Burenhult.
Oxford: ArchaeoPress (BAR Int. Series S1016)., pp. 463-468.
LOPEZ MOLINERO, A., CASTRO,A., PINO,J., PEREZ-ARANTEGUI, J.,
CASTILLO, J.R., 2000, “Classification of Ancient Roman Glazed Ceramics
using the neural network of self-organizing maps” Fresenius Journal of
Analytical Chemistry 367: 586-589.
Is it possible to design a
machine to do archaeology?
Fuzzy
Rules
Situated
explanation
Do we reject Artificial Intelligence
Tools and methods in Culture Heritage Research?
Or we can use it effectively!