MOBILE DECISION MAKING AND KNOWLEDGE
MANAGEMENT
Supporting Geoarchaeologists in the Field
Martin Blunn, Julie Cowie, David Cairns
Department of Computing Science and Mathematics, University of Stirling, Stirling, U.K.
Clare Wilson, Donald Davidson
School of Biological and Environmental Sciences, University of Stirling, Stirling, U.K.
Keywords: Mobile Decision Support, Archaeology.
Abstract: There is a professional responsibility placed upon archaeologists to record all possible information about a
given excavated site of which soil analysis is one important but frequently marginalised aspect. This paper
introduces SASSA (Soil Analysis Support System for Archaeologists), whose primary goal is to promote
the wider use of soil analysis techniques through a selection of ‘web based’ software tools. A description is
given of the field tool developed which supports both the recording of soil related archaeological data in a
comprehensive manner and provide a means of inferring information about the site under investigation.
Insight is gained through a user evaluating numerous decision trees relating to pertinent archaeological
questions. Whilst the field tool is capable of working in isolation, it offers a superior experience when
operated in unison with a Wiki. A brief discussion of the use of the Wiki application within the SASSA
project is presented.
1 INTRODUCTION
Applying geoarchaeology techniques to in-situ
monitoring of archaeological deposits can be an
arduous task for a non-specialist. They have to
identify questions relevant to the site and determine
suitable analytical techniques and sampling methods
to employ, all of which require complex decisions to
be made. Although there are many sources of
material citing good practice when analysing site
deposits, the nature of an archaeological dig does not
easily facilitate access to such information.
This paper introduces SASSA: a Soil Analysis
Support System for Archaeologists. SASSA is a
mobile knowledge-based tool that provides a means
by which pertinent information used to assist in
analysis is easily accessible yet provided in a
comprehensive manner. In addition, SASSA
incorporates decision support technologies to aid the
archaeologist in both their analysis of a site and
recording of important site data.
In Section 2 we begin by discussing the field of
geoarchaeology, highlighting some of the issues that
contribute to the complexity of the area. Section 3
reviews the current state of mobile decision support
research and outlines the relevance of SASSA to this
domain. Section 4 focuses on the SASSA field tool
with the underlying technology described in Section
4.1 and use of the tool discussed in Section 4.2.
Section 4.2 also illustrates how the tool might be
used by means of an example scenario. Conclusions
and future work are provided in Section 5.
2 GEOARCHAEOLOGY
Geoarchaeology is a broad discipline with much to
offer the archaeologist. Geoarchaeological research
has addressed a diverse number of issues; it has been
applied not only to archaeological sites and
landscapes, but also to experimental and
ethnographic research, and in-situ monitoring of
archaeological deposits.
57
Blunn M., Cowie J., Cairns D., Wilson C. and Davidson D. (2007).
MOBILE DECISION MAKING AND KNOWLEDGE MANAGEMENT - Supporting Geoarchaeologists in the Field.
In Proceedings of the Ninth International Conference on Enterprise Information Systems - AIDSS, pages 57-62
DOI: 10.5220/0002354500570062
Copyright
c
SciTePress
The archaeological questions addressed by soil
analysis fall broadly into two categories; site-
specific and landscape level. At the site level there
are issues over formation processes involving the
anthropogenic, depositional and human history of
the site. Such studies address the nature, source and
processes leading to the accumulation of deposits
through, for example, the use of multi-element
analysis to address space use and the identify
functional areas (Entwistle et al., 2000; Knudson et
al., 2004; Wells, 2004)
and the effects of
bioturbation
(Balek, 2002; Grave and Kealhofer,
1999) phosphatisation
(McCobb et al, 2003), and
waterlogging
(Caple, 1998) on the integrity of the
stratigraphic record. At the landscape level questions
concern human-environment interactions, for
example, the impact of human activity on erosion in
the wider landscape
(Wilkinson, 2005), resource
management in archaeological landscapes (Simpson
et al, 1998) and the effect of large scale natural
disasters on settlement location (Goff et al, 2003).
The complexity of the area means it can be
difficult for a non-specialist to identify the questions
relevant to a particular site and match this to the
relevant analytical techniques and sampling methods
or to critically evaluate the results. There is a good
geoarchaeological knowledge base already
available; it is the presentation of this material in a
manner that is easily accessible and comprehensible
to an interested non-specialist that is lacking. An
example of this is the use of multi-element
techniques to address questions of space use across
sites, or to identify sites within the landscape. The
question being asked influences the sampling
regime, and case studies of these two approaches
might include the identification of site extent at
Shapwick (Aston et al., 1998), or the identification
of activity areas in a classical site in Honduras
(Wells, 2004). English Heritage guidelines
(Avala et
al., 2004) provide questions associated with different
types of deposit linked to methods of investigation
and field diagnostic tools, such as finger texturing
flow charts, but access to specialist literature and the
time needed to absorb the specialised information
are still a problem for many archaeologists.
The development of the Soil Analysis Support
System for Archaeologists (SASSA) is aimed at
addressing the issues outlined above.
3 MOBILE KNOWLEDGE AND
DECISION SUPPORT
Knowledge management systems embrace
heterogeneous approaches for representing and
processing human knowledge to enhance decision-
making capability of human decision-makers (San
Pedro et al, 2003). With the event of mobile and
networked environments, there is a need to further
the approaches used. Mobile decision support and
knowledge systems have to handle the difficulties
and complexities brought about by context changes
in a mobile computing environment.
Mobile computing is a new technological
paradigm in which users access services via a range
of devices through a shared infrastructure, regardless
of their physical location or movement behaviour
(Zaslavsky
et al, 1998). Complexities and
uncertainties which derive from ensuring portability
of applications for a wide range of mobile devices
include frequent change in mode of operation, high
variability in performance and reliability, issues
surrounding visual display capabilities, finite
sources of energy, and facilitating recognition by the
system of the user, device and environment in which
the mobile computing takes place.
Related work in the area of mobile computer
support focuses on development of knowledge-based
services on hand-held computers (San Pedro et al,
2004; Cowie and Burstein, 2007). For example,
work on mobile clinical support systems, addresses
decision support such as knowledge delivery on
demand, medication consultant, therapy reminder
(Spreckelsen et al,
2000), preliminary clinical
assessment for classifying treatment categories(San
Pedro et al, 2003; Michalowski
et al 2003), and
providing alerts of potential drugs interactions and
active linking to relevant medical conditions (Chan,
2000). Most of these mobile support systems use
intelligent technologies and soft computing
methodologies (e.g., case-based reasoning, multi-
attribute utility theory) as background frameworks
for intelligent decision support.
This research builds on our existing work in the
area of mobile knowledge management, to provide a
central repository of archaeological information,
which is not restricted to location or platform. The
research addresses the changing way in which
information is required and decisions are made, the
impact this has on the type of systems developed,
and the emergent technologies that facilitate such
support. Our existing work in the area of developing
cross-platform systems designed for mobile /
(office) (Hodgkin et al, 2004; Cowie et al, 2006)
has
ICEIS 2007 - International Conference on Enterprise Information Systems
58
highlighted the suitability of such approaches to a
wide array of application areas. The SASSA field
tool extends this work both in the application area it
addresses and the technologies it uses.
4 THE FIELD TOOL
4.1 Technological Design
The field tool is tailored primarily towards two
categories of hardware device, static and mobile,
which in turn facilitate the use of the tool within
certain environments. The mobile category includes
devices such as a PDA (Personal Digital Assistant)
or ‘Smart phone’ whilst the static category
comprises of the more traditional desktop or laptop
computer. However, the two distinct functional
aspects of the field tool, recording the archaeological
data and evaluation of the data through the use of
decision trees, are both performed through the same
user interface and processing engine. The ubiquitous
HTML (Hyper Text Mark-up Language) browser
was identified as the most flexible interface for this
application, permitting user access via the hardware
devices listed above. The web interface technology
used is Sun Microsystem’s Java servlet technology
which is now common place within e-commerce
applications. Servlet technology was chosen due to
the availability of suitable API’s (Application
Program Interface), ease of design / implementation
of user session handling and its scalability under
load conditions. The Java servlet based system, used
in conjunction with the Apache Tomcat servlet
engine which assumes all conventional server
responsibilities, ensures that the tool utilises ‘open
source’ code and is thus freely available to all users.
Use of XML for data encoding
Through discussions with the domain specialists it
was determined that data storage requirements for
the project were not excessive and therefore both
user specified data and ‘knowledge data’ could be
stored within a proprietary XML (eXtensible Mark-
up Language) format data file. ‘Knowledge data’ is
defined within this tool as both descriptive content
and information associated with implementing the
tool. The XML format file is processed internally
using DOM (Document Object Model) methodology
to permit ease of manipulation and processing of
data. SASSA then produces output XML which is
rendered into XHTML through the use of XSLT
(eXtensible Style-sheet Language Translations)
technology. The output XML provides the structure
for the output content which is formatted for the
particular graphical user interface (GUI) of a given
user. The XSLT translation provides the flexibility
to tailor the style of presentation to the browser
being used. For example, the same output XML will
be used to generate different XHTML content
depending upon whether the user is accessing the
SASSA system via a PDA or desktop PC. It can also
be used to alter content to suit particular browsers or
even generate PDF equivalents of XHTML forms.
Figure 1: SASSA System Architecture.
A system diagram of SASSA is shown in
Figure 1 illustrating the principle software
components and their inter-relationships. The
normal process is for a user to log in to the SASSA
system at which point their XML data is loaded
into SASSA. A section of this XML data is then
requested, for example a ‘Site Details Form’. The
XML data encoding the site details is processed
into an intermediary XML output format
conveying more display related information. This
is then processed via an XSLT to produce
XHTML, and returned to the requesting
browser/device.
The SASSA field tool uses decision trees to aid
the geoarchaeologist in inferring information about
the site they are investigating. Insight is gained
through a user choosing from a compilation of
pertinent archaeological questions, each selected for
revealing the most information about the focus site
and ease / accuracy of answering. A decision tree is
formulated for each question comprising of a series
of sub-questions. It was felt the use of decision trees
(rather than other decision support techniques such
as case-based reasoning) was appropriate given that
feedback as to how an answer is attained is easy to
provide using this technique. In addition, the
infrastructure of decision trees seemed to be the
closest match to the current way in which decisions
XML
XHTML
Response
Request
HTTP
XML
User Data XML
XML
Parser / Builder
SASSA
Field Tool
Servlet
XSLT Parser
Configuration XML
PC XSL
PDA XSL
PDF XSL
Template XML
HTML
Browser
[PC/PDA/SmartPhone]
MOBILE DECISION MAKING AND KNOWLEDGE MANAGEMENT: Supporting Geoarchaeologists In The Field
59
about soil structures are made, thus facilitating an
easier transition from paper-based methods to the
SASSA tool than had a different approach been
adopted. A decision tree is constructed within a
proprietary XML data structure and evaluation
follows a standard ‘weighted sum’ methodology to
produce a specific ‘score’ for the question for a set
of specified answers to the sub-questions. The score
is presented to the user in a tabulated format that
contains not only the specific information on how to
interpret their score but that of other possible scores
in the answer range. This presentation format allows
the user to determine whether their results are
‘border line’ between score categories and thus draw
their own conclusions on the presented data. The
proprietary XML structure contains not only
presentation data (e.g. question text) and processing
data (e.g. weighting value for an answer) but allows
the implicit hierarchy of decision trees to be
captured through a combination of recursive data
structures and internal processing path links. A
similar hierarchical design structure is used for
storage of the site information archive and therefore
a common ‘XML processor’ has been used to
interpret both data structures.
4.2 Use of the SASSA Field Tool
When recording the archaeological record of a new
excavation two key aspects from a soil science
perspective have been identified; consistently
recording relevant data across multiple samples and
providing answers to a range of pertinent questions
which reveal more general information about a
particular site. Both of these functions have been
incorporated into the SASSA field tool. It is
envisaged that SASSA will be used both ‘on site’ for
immediate help with excavation decisions or in an
office environment when analysing retrieved data.
To facilitate this, SASSA is operable on a variety of
hardware platforms, each tailored to permit ease of
use within different situations. For example,
operating the field tool on a PDA (personal digital
assistant) or ‘Smart Phone’ will allow recording of
data within the excavation trench of interest.
Operation on a laptop/desktop PC within a site
office can facilitate group discussion on recorded
results at the end of a day’s recording activities.
4.2.1 Example Use of SASSA
Recording site information
Prior to a team of archaeologists arriving at an
excavation site, a significant amount of preparation
work and information gathering has been performed.
Some of this data is pertinent to the soil record of the
site and is important to document. Site information
is recorded in a hierarchical manner, where data can
relate to the site itself, specific sections or trenches
within a site, and contexts or layers within each
section. In accordance with this hierarchy, the
SASSA tool provides the facility to enter data about
all aspects of a site as appropriate. In the screenshot
shown in Figure 2, we can see the user is providing
colour information for the Site Castle / Main Trench
Figure 2 : SASSA Site Details.
Figure 3: Field Tool Interface.
ICEIS 2007 - International Conference on Enterprise Information Systems
60
/ Top Layer. The user is supported in this data
collection activity with a clear, ‘form style’ format
of data entry and links to textual definitions of
unusual terms and other information sources (such
as diagrams, photographs and case studies).
Use of decision trees for soil context discovery
The SASSA field tool uses decision trees as a means
of identifying the likelihood of certain soil properties
being present at a given site. Currently, the tool
focuses on what we regard as four of the most
prominent properties investigated at a site: the
presence of burnt material, whether the burnt
material (if present) occurred in-situ, the presence of
buried soil, and whether the deposit is natural. By
walking the user round the decision tree associated
with each of these properties, the tool collates
knowledge of the site and provides inferences
concerning the soil context. Figure 3 shows an
example of the questions the user is asked in relation
to the issue of burnt material. The questions asked,
and how the information provided by the user, is
utilised in providing new information to the user
dictated by the decision tree associated with the
issue being explored. In the example shown, the
answers supplied by the user provide an evaluation
score of 35. As is depicted at the bottom of Figure 3,
score ranges are provided along with how such
scores should be interpreted.
In addition to providing inferences about the site,
suggestions for further routes of investigation are
given. Such investigations include suggestions for
further tests that could be conducted and links to
further material provided by the SASSA tool that
could provide relevant case study or background
information.
5 CURRENT AND FUTURE
WORK
SASSA Extensions
The current focus of our work involves development
of the main SASSA website as a Wiki, from which
the field tool can be accessed. The ‘Wiki’ class of
software tool, typified by the world’s largest on-line
encyclopaedia ‘Wikipedia’ (Aronsson, 2002) allows
browser displayed content to be uploaded and
modified by registered users rather than just the site
developer. This ensures that content can remain
accurate and applicable as well as evolve with the
subject over a period of time. ‘MediaWiki’, has been
adopted for the SASSA system since it is the
technology on which ‘Wikipedia’ is based and users
are therefore more likely to be familiar with it.
SASSA Evaluation
Acceptance of the SASSA tools by the intended user
community is critical to the success of this project.
At the current stage of development, most effort has
been assigned to developing prototype versions of
the field tool. Evaluation of these development
models thus far has taken place through informal
discussion with a small selection of interested
persons. It is the intent however, to evaluate the
prototype tools more thoroughly at a series of formal
sessions across the UK at regular intervals in the
remaining time of the project.
Application to other domains
The infrastructure of SASSA has been developed
such that it is independent of the application,
ensuring portability of the system. We hope to
investigate other areas where the tool might be
applied once all components of the software are
complete.
ACKNOWLEDGEMENTS
We gratefully acknowledge the funding received for
this research by the Natural Environment Research
Council (NE/D000971/1).
REFERENCES
Aronsson, L 2002. Operation of a large scale general
purpose Wiki website: experience from susning.nu’s
first nine months in service. In Proceedings of the 6th
International ICCC/IFIP Conference on Electronic
Publishing, Czech Republic, November 6-8, 2002.
Verlag für Wissenschaft und Forschung.
Aston, M.A 1998. Chemosphere, 37; 465-477.
Ayala, G., Canti, M., Heathcote, J., Sidell, J., and Usai, R.,
2004. Geoarchaeology: using earth sciences to
understand the archaeological record. English
Heritage.
Balek, C. L. 2002. Buried artifacts in stable upland sites
and the role of bioturbation: a review. Geoarchaeology
17: 41-51.
Caple, C. 1998. Parameters for monitoring anoxic
environments. In (Eds.) M. Corfield, P. Hinton, T.
Nixon & M. Pollard, Preserving archaeological
remains in-situ. MoLAS, London: 66-72
Chan, A. 2000. WWW+ smart card: Towards a Mobile
Health Care Management System, International
Journal of Medical Informatics, 57, 127-137.
MOBILE DECISION MAKING AND KNOWLEDGE MANAGEMENT: Supporting Geoarchaeologists In The Field
61
Cowie, J., Burstein, F. 2007. Quality of data model for
supporting mobile decision making. Journal of
Decision Support Systems, In Press
doi:10.1016/j.dss.2006.09.010.
Entwistle, J.A., Dodgshon, R.A. and Abrahams, P.W.
2000. The geoarchaeological significance and spatial
variability of physical and chemical soil properties
from a former habitation site, Isle of Skye. Journal of
Archaeological Science 27: 287-303.
Goff, J.R. and McFadgen, B.G. 2003. Large earthquakes
and the abandonment of prehistoric coastal settlements
in 15th Century New Zealand. Geoarchaeology 18:
609-623.
Grave, P. & Kealhofer, L. 1999. Assessing bioturbation in
archaeological sediments using soil morphology and
phytolith analysis. Journal of Archaeological Science
26: 1239-1248.
Hodgkin, J., San Pedro, J., Burstein, F. 2004. Quality of
Data Model For Supporting Mobile Decision Making.
In Proceedings of Decision Support Systems
Conference, July 1-3, Italy.
Knudson, K.J., Frink, L., Hoffman, B.W. and Price, T.D.
2004. Chemical characterization of Arctic soils:
activity area analysis in contemporary Yup’ik fish
camps using ICP AES. Journal of Archaeological
Science 31: 443-456.
McCobb L.M.E., Briggs, D.E.G., Carruthers, W.J. and
Evershed, R.P 2003. Phosphatisation of seeds and
roots in a Late Bronze Age deposit at Potterne,
Wiltshire, UK. Journal of Archaeological Science 30:
1269-1281.
Michalowski, W., Rubin, S., Slowinski, R.,Wilk, S. 2003.
Mobile Clinical Support System for Pediatric
Emergencies, Decision Support Systems, 36, 161-176.
San Pedro, J. and Burstein, F. 2003. An intelligent
decision support model for tropical cyclone prediction,
submitted to International Journal of Intelligent
Systems in Accounting, Finance and Management,
Special Issue on Decision Making and Decision
Support.
San Pedro, J., Burstein, F., Churilov, L., Wassertheil, J.,
Cao, P. 2003. Intelligent Multiattribute Decision
Support Model For Triage. In Proc. of International
Conference in Information Processing and
Management of Uncertainty in Knowledge-Base
Systems, July 4-9, 2004, Perugia, Italy.
San Pedro, J., Burstein, F., Zaslavsky, A., Hodgkin, J.
2004. Pay by Cash, Credit or EFTPOS?
Supporting the User with Mobile Accounts Manager.
In Proc. of the 3rd Mobile Business Conference,
MBusiness Conference, New York, USA, 12th - 13th
July, 2004.
Simpson, I.A., Bryant, R.G. and Tveraabak, U. 1998.
Journal of Archaeological Science 25: 1185-1198.
Spreckelsen, C. Lethen, C., Heeskens, I., Pfeil, K.,
Spitzer, K. 2000. The Roles Of An Intelligent Mobile
Decision Support System In The Clinical
Workflow,URL
www.citeseer.nj.nec.com/spreckelsen00roles.html,
Accessed 11 Nov 2003.
Wells, E.C. 2004. Investigating activity patterns in
Prehispanic plazas: weak acid-extraction ICP-AES
analysis of anthrosols at Classic period El Coyote,
Northwestern Honduras. Archaeometry 46: 67-84.
Wilkinson, T.J. 2005. Soil erosion and valley fills in the
Yemen highlands and southern Turkey: integrating
settlement, geoarchaeology and climate change.
Geoarchaeology 20: 169-192.
Zaslavsky, A. and Tari, Z. 1998. Mobile computing:
Overview and current status. Special issue of
Australian Computer Journal on Mobile Computing,
30, 2, 42-52.
ICEIS 2007 - International Conference on Enterprise Information Systems
62