INTEGRATION OF SPATIAL TECHNOLOGIES AND

SEMANTIC WEB TECHNOLOGIES FOR INDUSTRIAL

ARCHAEOLOGY

Ashish Karmacharya

1,2

, Christophe Cruz

, Frank Boochs

and Franck Marzani

Institut i3mainz, am Fachbereich 1 - Geoinformatik und Vermessung Fachhochschule Mainz

Holzstrasse 36, 55116 Mainz, Germany

Laboratoire Le2i, UMR-5158 CNRS, UFR Sciences et Techniques, Université de Bourgogne

B.P. 47870, 21078 Dijon Cedex, France

Keywords: OWL, SWRL, Spatial analysis, Knowledge management, Ontology, Semantic web, Industrial archaeology.

Abstract: We propose a method that uses the advancement in spatial technologies from current database systems

within the Semantic Web Technologies in order to enrich and to populate the knowledge of a domain

defined in an OWL-DL ontology. The results of spatial operations and functions are used to populate and to

enrich ontologies with new individuals and new relationships. The advantage of spatial analysis within

Semantic Web technologies is the diversity of the functionalities provided by the combination of spatial

operations and the rule language of the Semantic Web (SWRL). This method is applied in the industrial

archaeology domain in order to enhance the knowledge management.

1 INTRODUCTION

Geometry has always been the dominant component

in any system related to an archaeological project.

The objects extracted on the excavation sites are

represented by using their geometries. This fact has

led to the assumption that a system related to such

projects is either a 3D object modeling system or

Geographic Information System (GIS), as they both

use object geometries and their relations with their

surroundings. However, in the whole process the

semantics of the geometric objects and their

relationships with the surroundings are neglected.

With the advancement of survey technologies, data

can be collected more accurately. On the one hand,

this has brought a great advantage in the analysis

process as we possess more and diverse data to

perform the precise analysis. On the other hand, it

has created difficulties in managing them with

existing database systems due to their size and

diversity. This issue is even more visible in an

industrial archaeology project. Indeed, the sites of

excavations are available for a very limited time

only and thus the data have to be collected and

stored in a very short time. In addition, the diversity

of the data makes the management of the

information with the existing database systems more

complex. Hence, a lot of research is done in the field

of data indexation and information retrieval in order

to reach the level where this vast amount of

information can be managed through the knowledge

defined by the archaeologists. Actually, the

knowledge about the objects excavated from the

sites can only be defined by the archaeologists.

Consequently we propose a method which is

adjusting the old methods while, at the same time,

taking advantage of the emerging cutting edge

technology. We propose in our method to retain the

storing mechanism with the existing database

management systems and to consider geometry as

one of the major data types. In addition we suggest

the use of a collaborative web platform based on

semantic web technologies and knowledge

management so that the information can be handled

by several archaeologists and technicians. The

platform will allow to store data during the

excavation and to manage it through the knowledge

acquired during the identification process.

Furthermore, it facilitates the collaborative process

between the archaeologists concerning the

generation of knowledge from the data sets. The

Karmacharya A., Cruz C., Boochs F. and Marzani F.

INTEGRATION OF SPATIAL TECHNOLOGIES AND SEMANTIC WEB TECHNOLOGIES FOR INDUSTRIAL ARCHAEOLOGY.

DOI: 10.5220/0002791300750080

In Proceedings of the 6th International Conference on Web Information Systems and Technology (WEBIST 2010), page

ISBN: 978-989-674-025-2

main principle of our approach is the use of semantic

annotation to provide a semantic view on the data

sets. The shared ontology that defines an index on

the semantic annotations allows us to build a global

schema between the data sources. This global

schema allows us to annotate, index, search and

retrieve data and documents.

The semantic tool is being used in a wide range of

applications ranging from data integrations to

knowledge management. Given that, this is a

relatively new topic, so great amount of researches

have been conducted on the different aspects of this

technology. However, most of the research hardly

includes spatial information and if they do they are

primarily focusing on spatial data integration with

semantic technologies (Green, 2008). The

ArchaeoKM (Karmacharya, 2008) project aims at

the inclusion of the spatial data process within

Semantic Web technologies in order to not only

establish a comprehensive data integration process

between spatial data but to also combine the benefits

of spatial operations with the deductive reasoning

capabilities of OWL DL ontologies for a

comprehensive knowledge management. The benefit

of spatial analysis within Semantic Web

technologies lies in the diversity of the

functionalities provided by the combination of the

spatial operation and the rule language of the

Semantic Web.

In the following section, we will discuss the

technical background of the project. In section 3 we

will introduce the Web platform ArchaeoKM.

Section 4 focuses on the spatial facilitator. It

explains the spatial integration of functions and

operations concerning the enrichment of ontologies,

as well as the SWRL extension. The last section

concludes the paper.

2 BACKGROUND

The sharing of knowledge in archaeology and it

disseminate to the general public through wiki has

been discussed in (Costa, 2008). Likewise the use of

knowledge to build up a common semantic

framework has been discussed in (Kansa, 2008).

Research works exist in the field of archaeology, but

most of the research is carried out in other related

fields. The existing research focusses more on the

use of a common language for efficient

interoperability. The research project in (Kollias,

2008) concerns the achievement of syntactic and

semantic interoperability through ontologies and the

RDF framework in order to build a common

standard. Data integration through ontologies and

their relationships are discussed in (Doerr, 2008).

Although the work on semantic web and knowledge

management in the field of Information systems in

archaeology or related fields has made progress with

these research works, it remains a fact that they are

in a very preliminary phase today. In addition, these

projects concentrate more on how to achieve

interoperability with semantic frameworks and

ontologies. However, no research focuses on the

knowledge generation process and more specifically

on rules defined by archaeologists in order to build

up the system which will use, evaluate and represent

the knowledge of the archaeologists.

Industrial Archaeology is perhaps the best suited

field in archaeology on which to carry out our

research. Actually, Industrial Archaeological Sites

(IAS) are available for a very short time only. The

limited time available for the storage of the data is

one of the concerns we want to address here.

Moreover, the amount of data that has to be

collected in this short span is very large and diverse.

The ArchaeoKM project focuses their attention on

the site of the Krupp factory in Essen, Germany. The

200 hectares area was used for steel production

during the early nineteenth century and was

destroyed in the Second World War. Most of the

area has never been rebuilt and thus provides an

ideal site for industrial archaeological excavation.

The area will be used as a park of the ThyssenKrupp

main building in 2010. Actually, we are running out

of time as far as the collection of the data is

concerned. The first challenge consists in creating a

relevant data structure which helps to retrieve those

data efficiently. In addition, the amount of data that

has to be collected is huge, so the system has to be

able to handle a huge data set.

The nature of the data set generated during the

project is heterogeneous. As it can be seen, the

acquired data ranges from scanned point clouds from

terrestrial laser scanners to the floor plans of old

archives. The primary source of geometric

information is provided through a point cloud. The

point clouds have a resolution of 0.036 degrees and

are in the Gauss Krüger coordinate system, zone II

(GK II). This is the main data set used for the 3D

object modeling. Beside the point clouds, a great

amount of images are also collected during the

excavation. Most of the images are taken with a non

calibrated digital camera and, consequently, do not

contain any information about the referencing

system. Even though they do not contain any

referencing information, they posses vital semantic

information and can be used for the formulation of

WEBIST 2010 - 6th International Conference on Web Information Systems and Technologies

knowledge. However, photogrammetric flights take

place to acquire aerial images of the area. The aerial

images are processed to generate a digital

orthophoto with a resolution of 10 cm. The digital

orthophoto is again in the Gauss Krüger referencing

system (GK II). Furthermore, a huge amount of

archive data are collected. These data contain floor

plans, old pictures and other semantic information.

Likewise, the notes taken by archaeologists are also

important to acquire semantic information of the

findings. ArcGIS databases are also available

depending on the site and its nature. These databases

are in the GK II reference system. For our example,

this database gives an overview of the site and can

be overlayed with the orthophoto in order to identify

the interesting locations easily.

3 THE ARCHAEOKM

PLATFORM

ArchaeoKM is a Web platform which takes into

consideration an adjustment of the old methods and,

at the same time, takes advantage of the emerging

cutting edge technology. The system still proposes

to retain the storing mechanism with the existing

database management systems and to consider

geometry as one of the major data types. In addition,

we suggest the use of a collaborative Web platform

based on semantic web technologies (OWL, RDF,

SPARQL, SWRL) and knowledge management in

order to handle the information provided by several

archaeologists and technicians. ArchaeoKM includes

deductive rules defined by archaeologists on data of

excavated objects. The knowledge is stored in a

machine-readable format. Consequently, the

knowledge can be translated into a human-readable

format.

The Web based system ArchaeoKM has an

architecture divided into three major levels. Each

level has its own distinct functionality and is

interdependent with the others. The syntactic level

stores all the information that is excavated on the

site. As discussed earlier, information is either stored

in file formats like images or archive data or stored

in the Relational Database Management System like

archaeological notes or scanned/GIS data.

The semantic level allows the management of

generated knowledge. It is achieved through the

ontological structure set up by archaeologists.

Archaeologists are involved actively in this phase as

they are the ones best suited to provide entities and

their relationships needed to build up the domain

ontology. This level represents a bridge between

interpretative semantics in which users interpret

terms and operational semantics in which computers

handle symbols (Guarino, 1994).

The knowledge level represents the specification

of the knowledge of archaeologists concerning the

industrial findings. This level provides the user with

a graphical interface represented by Web pages in

order to display the generated knowledge. The pages

are interrelated and can be navigated according to

their relevancy.

Besides these three levels, the system architecture

contains a component that facilitates the knowledge

generation, update and validation through a spatial

perspective. This component called the “spatial

facilitator” is in charge of the spatial data analyses

and provides thus the result in order to enrich and to

populate the ontology. The ontology population

process is the activity of adding new instances to an

ontology. The ontology enrichment is the activity of

extending an ontology by adding new elements (e.g.

concepts, relations, properties, axioms) (Castano,

2007). The details of the component are given in the

next section.

4 THE SPATIAL FACILITATOR

This section highlights our approach to the

management of the spatial operations in order to

enrich and to populate our ontology. The ontology

schema of the ArchaeoKM platform is responsible

for maintaining a relation between the enrichment of

the ontology, with the corresponding individuals

which are the objects excavated from the site, and

their semantic annotations on the data and

documents. The ontology schema is also responsible

for reflecting the archaeological interpretations of

the objects through proper relationships between

different entities of the objects.

4.1 The Ontology Schema

The core of the schema is the concept “siteFeature”

which stores all the excavated objects. The basic

process behind the “ArchaeoKM” is very

straightforward. Archaeologists are responsible for

the indexation of the findings on the orthophoto.

Those findings are then enriched in the domain

ontology through respective objects. The spatial

facilitator covers also the adjustments carried out

within the ontology schema in order to incorporate

the spatial components. The ontology schema

represents the terminological definition. It is defined

INTEGRATION OF SPATIAL TECHNOLOGIES AND SEMANTIC WEB TECHNOLOGIES FOR INDUSTRIAL

ARCHAEOLOGY

with the OWL-DL language which is a description

logic language (Baader, 2003). Actually, it

represents the definition of concepts and roles which

are properties and relations between concepts.

The ontology schema in the ArchaeoKM platform

has to be adjusted in order to incorporate the spatial

functions and operations. In general, the spatial

operations and functions provided by the current

database system can be broadly categorized into two

categories – spatial processing functions and spatial

relationship functions. The first category represents

unary functions and the second represents binary

functions. The unary operations return the new

geometry itself whereas the binary operations return

the binary value. Figure 1 shows the two categories

of the spatial functions. By adding spatial relations

between site features (feat:siteFeature), archeologists

define a certain kind of knowledge concerning the

disposition of findings on the current site. For

instance, a knowledge specification about the

domain can be made as it exist a finding “oven” and

a finding “railway” that overlap a finding

“building”. It means that the building is a finding

“factory”. So, a concept “feat:factory” is defined as

a subclass of “feat:siteFeature” and with the

condition described previously. It can be easily

computed with the help of the 2D/3D annotations of

the indexes.

4.2 Enrichment of the Ontology

Schema by Adding Spatial

Operations

The two sets of spatial operations are represented

with two different approaches in the ontology

schema. The first set of spatial operations needs to

be treated as we treat the features excavated in the

concept “siteFeature” since they result in

geometries. This is achieved by introducing a new

concept “spatialAnalysis” with sub-concepts to

support such 2D and 3D operations. It is important

to define a property that represents the relationship

between the spatial concepts with the feature

excavated. It is defined through predicate

“hasSpatialAnalysis”.

The second set of spatial operations provides the

status of the particular relationship between two

objects. Such relationships are binary relationship

and they show whether there exists a particular

relationship between two objects or not. As these

relationships do not yield new geometry and they

perform much in similar line to the object

relationships, they are represented as a form of

object relationship. It is shown by

“hasSpatialRelAnalysis” and has both range and

domain as “siteFeature”. It is possible to perform

binary spatial operations between the objects of

“siteFeature” and “spatialAnalysis”. From this point,

it can be see that spatial information which defines

the knowledge of a domain can be added in the

ontology. In addition the properties and relationships

can be verified with the help of spatial database by

the spatial facilitator.

Figure 1: Two types of spatial operations (a) Buffering

(spatial processing) a linear feature (red linear feature)

generates crossed polygonal features around it (b) Five

polygons to demonstrate the touch (Spatial Relationship)

options – A touches B Æ true, A touches D Æ false.

In order to define the new spatial relationships

between individuals, any individual from the

concept “feat:siteFeature” has relationship to an

individual “shape:Feature” which can be 2D or a 3D

shape. Almost all of the existing database system

supports storage and retrieval of the spatial data with

their spatial extensions. They also support spatial

operations on these data. However the scales of

spatial operations vary from one database system to

another. They also vary in the support for the 3D

data set. Currently, there are not many 3D spatial

operations supported by the existing database

systems. Oracle 11g (Oracle, 2007) and PostGIS

1.3.5 (PostgresSQL, 2008) of PostgreSQL 8.3 are

the leading database systems supporting the 3D

operations. However such operations are mostly

limited to unary operations. ArchaeoKM intends to

use the advancements in spatial operations in

PostGIS to enrich the ontology. All the operations

are carried out in accordance to the SQL syntax of

the spatial operations of the database systems and

will be performed on the data stored in the database.

The results that are generated through such

operations are used to enrich the ontology. In this

manner, the database is merely used as the tool to

store the spatial data and to carry out the required

WEBIST 2010 - 6th International Conference on Web Information Systems and Technologies

spatial operations. The relationships and the results

are managed through the ontology.

4.3 The Extension of SWRL with

Spatial Analysis

This section presents the method used to integrate

the spatial operations (unary, binary, 2D and 3D

operations) with the help of Horn clauses (SWRL

language) in order to define knowledge on the

Industrial Archaeological Site (IAS) with ontologies

as well as rules.

Example of a SWRL expression. The following

example creates a new relationship

“cooperatedWith” between authors if they worked

on the same publication.

Publication(?a) ∧

hasAuthor(?a, ?y) ∧

hasAuthor(?a, ?z) ∧

differentFrom(?y, ?z)

→ cooperatedWith(?y, ?z)

In addition it exists “built-in” predicates in the

SWRL language that allow the computation of

advance information. For instance: Person(?p) ^

hasAge(?p, ?age) ^ swrlb:greaterThan(?age, 18) →

Adult(?p). The method presented in this section

consists in showing how can be extended the built-in

predicates with spatial operation.

Examples of a SWRL rule (Horn clauses) with a

spatial operation Buffer. The result is a new relation

between site feature x and y that satisfy the “buffer”

operation. Actually it consists to enrich the ontology

by adding a new relation between two individuals.

river(?x) ∧

Building(?y) ∧

archaeokm:Buffer(?x, ?y, 50)

→ isLiableToFlooding (?y)

oven(?x) ∧

Building(?y) ∧

archaeokm:Buffer(?x, ?y, 50)

→ hasOven(?y, ?x)

The spatial facilitator component in our

architecture is composed of an SWRL engine

improved with spatial operation predicates that

allow the definition of complex rules. In order to

realize these operations, each spatial operation is

converted to an SQL request. The example buffer

given is a combination of spatial operation Buffer

and “within”. The operation consists to define a SQL

query (e.g. next example). Thus, an SQL query has

to be defined for every SWRL operation in order to

process the expected result.

Example of a built-in operation converted into an

SQL request

SELECT y

FROM Ovenbb_tb

WHERE

within(the_geom,

buffer((

SELECT the_geom

FROM Ovenbb_tb

WHERE name = x),50)

)

4.4 An Example

This section presents a scenario to present our case

study. We are using the bounding boxes of five

distinct objects that are found in the industrial

archaeological site. Those findings are specialized

concepts of the concept “siteFeature”. These

concepts are “Oven”, “Railway”, “Structure”,

“Chimney”, “Pipeline” and “Plant”. Those concepts

represent the objects excavated.

Once the objects are excavated from the site, they

are used to enrich the ontology against their

respective concept. The geometries of these objects

are stored in the PostgreSQL database as the spatial

data type provided by PostGIS – spatial extension of

the database system.

To illustrate the spatial operations we discussed

in the previous sections, we take one operation from

each unary and binary spatial operation and

demonstrate how they enrich the ontology. To begin

with we take “Buffer operation” which buffers the

feature and is a unary operation. We define a buffer

of 50 meters around the “Oven_1” and populate the

ontology with a new specialized concept “Buffer” of

the concept “spatialAnalysis”. Then we populate this

concept with the corresponding object

“buffOven_1_50m” and store the resulted

coordinate.

Example of a spatial operation “buffer”.

SELECT AsText(

buffer((

SELECT the_geom

FROM Ovenbb_tb

WHERE name = 'Oven_1') ,50))

It is clear that when we specialize the concept

“spatialAnalysis”, a respective specialized object

property under “hasSpatialAnalysis” has to be

created too. In this case “hasBuffer” has to be

created under “hasSpatialAnalysis” simultaneously.

So the new RDF triplet from the operation above

would be (“siteFeature”, “hasBuffer”, “Buffer”).

The knowledge base is then populated with

“Oven_1” “hasBuffer” “buffOven_1_50m”.

INTEGRATION OF SPATIAL TECHNOLOGIES AND SEMANTIC WEB TECHNOLOGIES FOR INDUSTRIAL

ARCHAEOLOGY

The next operation is the binary operation and we

take as an example the “within” operation which

will show whether or not an object is contained in

the next one. This will generate binary results of the

operations. But to make this operation more

appropriate for our case, we modify the operation so

that it will extract all objects within the feature. The

spatial operation listed below will list out all the

features that are within the feature “Plant_1”.

Example of a spatial operation “within”.

SELECT name

FROM Ovenbb_tb

WHERE

within( the_geom,

(SELECT the_geom

FROM Ovenbb_tb

WHERE name = 'Plant_1')

))

The binary operations are used as the object

property “hasSpatialRelAnalysis” in the ontology. A

new specialized property “hasWithin” is created

with the RDF triplet as (“siteFeature”, “hasWithin”,

“siteFeature”). The knowledge base is then enriched

with these triplets (“Plant_1”, “hasWithin”,

(“Oven_1”, “Railway_1”, “Pipeline_1”, etc.)).

5 CONCLUSIONS

In this paper, the ArchaeoKM platform has been

presented by focusing on spatial analyses and by

showing the combination of these analyses with

Semantic Web technologies. These benefits are

materialized by the population and the enrichment

processes of a domain ontology with the help of

spatial operations using industrial archaeological site

data. An additional benefit is the extension of the

SWRL language by providing built-in “spatial

operations”. Thus, this extension allows the

definition of rules supplying new knowledge on the

IAS. These processes are managed by the spatial

facilitator component of the ArchaeoKM platform.

Although the case study uses industrial archaeology

for the description of the approach, it can be used in

other areas where the spatial data are the

predominant data type. Future work will be the

identification of all spatial operations that can be

handled by spatial database systems in order to offer

an overview of its capabilities. At the moment only

few of them are prototyped as a proof of concept.

REFERENCES

Baader, F., et al, 2003, The Description Logic Handbook –

Theory, Implementation and Applications, ISBN:

0521781760, Cambridge University Press, January

Castano, S., Espinosa, S., Ferrara, A., Karkaletsis, V.,

Kaya, a., Melzer, S., Moller, R., Montanelli S.,

Petasis, G., 2007. Ontology Dynamics with

Multimedia Information: The BOEMIE Evolution

Methodology, In Proc. of International Workshop on

Ontology Dynamics (IWOD) ESWC 2007 Workshop,

Innsbruck, Austria

Costa, S. and Zanini, E., 2008. Sharing knowledge in

archaeology: looking forward the next decade.

Digital Heritage in the New Knowledge Environment:

Shared spaces & open paths to cultural content, 31

October – 02 November in Athens, Greece

Doerr, M., 2008. The CIDOC Conceptual Reference

Model – A New Standard for Interoperability. Digital

Heritage in the New Knowledge Environment: Shared

spaces & open paths to cultural content, in Athens,

Greece

Green, J., Dolbear, C., Goodwin, J., 2008. Creating a

semantic Integration System using Spatial Data, 7th

International Semantic Web Conference (ISWC2008),

Karlshue, Germany, October 26-30

Guarino, N., 1994; The ontological level, in R. Casati B.

S. & White G., eds, Philosophy and the cognitive

sciences, Hölder-Pichler-Tempsky

Kansa, E. C., 2008. Opening Archaeology to Mash-ups:

Field Data and an Incremental Approach to Semantic,

Digital Heritage in the New Knowledge Environment:

Shared spaces & open paths to cultural content in

Athens, Greece

Karmacharya, A., Cruz, C., Marzani, F., Boochs, F., 2008.

Industrial Archaeology: Case study of Knowledge

Management for Spatial Data of Findings. 2nd

International Workshop on Personalized Access to

Cultural Heritage, in conjunction with 5th

International Conference on Adaptive Hypermedia and

Adaptive Web-Based Systems, Hannover, Germany

Kollias, S., 2008. Achieving Semantic Interoperability in

Europeana, Digital Heritage in the New Knowledge

Environment: Shared spaces & open paths to cultural

content, 31 October – 02 November in Athens, Greece

Oracle, 2007. Oracle Spatial Developer Guide 11g

Release, Oracle

PostgreSQL, 2008. PostGIS Manual, PostgreSQL

documentation

WEBIST 2010 - 6th International Conference on Web Information Systems and Technologies