KNOWLEDGE REPRESENTATION IN ENVIRONMENTAL IMPACT

ASSESSMENT

A Case of Study with High Level Requirements in Validation

Juli´an Garrido and Ignacio Requena

Department Computer Sciences and Artiﬁcial Intelligence, Granada University

Daniel Saucedo Aranda s/n 18071, Granada, Spain

Keywords:

EIA, Environmental Impact, Knowledge Representation, Ontology, Ontology validation, OWL.

Abstract:

An ontology which allows to represent knowledge of environmental impact assessment (EIA) and its involved

processes is presented. This knowledge representation is designed to be used in two contexts. The ﬁrst one

is a repository of the elements and concepts used in EIA by environmental experts as a structured knowledge

source. The second one is a formal deﬁnition of the concepts to be used in an intelligent system for EIA. The

ﬁrst usage requires obtaining a high level of consensus about what elements have to be deﬁned.

1 INTRODUCTION

Environmental impact assessment (EIA) ensures that

environmental consequences of projects are identi-

ﬁed, assessed and taken into account before the de-

cisions are made. A lot of methodologies and tools

for EIA have been built although most of them were

designed for a speciﬁc human activity. In fact, soft-

ware based on word computing and methodologies

for mining, civil construction and landﬁlls have been

developed by our group (Delgado et al., 2006).

These experiences have encouraged us to develop

a generic tool for EIA which is able to assess any in-

dustrial activity. This ambitious project requires more

sophisticated approaches due to the complexity. Con-

sequently, processes and elements involved in EIA

have to be gathered and organized using knowledge

representation by mean of ontologies.

The ﬁnal software will be based on this knowl-

edge representation and the ontology is expected to be

used by environmental experts as a knowledge source

so that a consensus in its contents must be reached.

Approaches for validation already exist although they

are focused on metrics to design a formal model for

evaluation (Gangemi et al., 2006; Brank et al., 2005)

whereas this ontology needs a higher effort in evalu-

ating their contents.

The ﬁrst section describes the methodology fol-

lowed to build the ontology. Next section shows

the most signiﬁcant design decisions and the general

structure. Section 4 explains a developed web appli-

cation to manage the ontology evaluation to reach a

consensual knowledge. And ﬁnally further work and

references are included.

2 METHODOLOGY

Nowadays, there is no standard deﬁned for building

ontologies so that each developer tends to follow his

own design criteria and principles depending on the

ontology usage.

However, some authors suggest guidelines and

methodologies such as the Uschol and King’s method

(Uschold and Gr¨uninger, 1996), the method used in

the KACTUS project (Schreiber et al., 1995), the Sen-

sus project (Swartout et al., 1997) or the Methontol-

ogy method (Fern´andez et al., 1999).

In (G´omez and Benjamins, 1999) it is explained

that most of methods do not put enough attention

in development phases like management, merging,

learning, integration, tracing and evaluation whereas

they focus its effort on conceptualization and ontol-

ogy implementation.

Combining the previous methodologies explained

above, we have established a set of steps to build the

proposal of EIA ontology.

1. The aim and scope identiﬁcation.

2. The ontology construction process.

(a) Getting knowledge: Top-down.

(b) Encoding phase.

412

Garrido J. and Requena I. (2009).

KNOWLEDGE REPRESENTATION IN ENVIRONMENTAL IMPACT ASSESSMENT - A Case of Study with High Level Requirements in Validation.

In Proceedings of the International Conference on Knowledge Engineering and Ontology Development, pages 412-415

DOI: 10.5220/0002295804120415

 SciTePress

(d) Inference.

3. Evaluation.

4. Documentation.

5. Maintenance.

In the following paragraphs, the most signiﬁcant

points of the ontology-building steps are brieﬂy com-

mented below.

The aim and the scope consist of building an on-

tology which establishes a conceptual framework for

EIA with two main objectives. On the one hand, it

must provide a structured knowledge which can be

used by environmental experts as a reference guide.

Moreover, it can be a helpful tool during the new

methodologies development for EIA. On the other

hand, the ontology cannot be only used as a glossary

for environmental experts. It includes formal seman-

tic deﬁnitions for processing and reasoning tasks.

The ontology has been designed with a top-down

strategy because new knowledge can be aggregated in

the future. This new knowledge will correspond with

concepts belonging to the EIA of activities not still

considered.

A strict and complex getting knowledge phase

must be accomplished due to the ontology will be

used by environmental experts. Hence they must par-

ticipate in the evaluation phase too.

Prot`eg`e

has been chosen for building the on-

tology due to its characteristics. Firstly, it per-

mits ontologies creation in OWL

, avoiding the di-

rect treatment of OWL syntaxes and following the

W3C (World Wide Web Consortium) recommenda-

tions. Secondly, it is an open-source platform, so it

can be customized to create knowledge models and

data inputs, e.g. adding vagueness or uncertainty to

the concepts, or using plugins. Furthermore, it works

with external dig reasoners like Pellet

or Racer

A lot of ontologies related to the environmental

sciences have been found, although none of them re-

lated to EIA. Most of these ontologies were rejected

because of an unsuitable granularity or content, so

that only a few existing concepts have been included.

The evaluation process has been accomplished

during the whole development in a cyclic way. Ex-

perts of environmental technologies area have evalu-

ated prototypes with different granularity levels and

they have provided a general review and new knowl-

edge.

Developed by Stanford Medical Informatics.

OWL (Ontology Web Language) is a semantic markup

language for publishing and sharing ontologies.

http://pellet.owldl.com/

http://www.racer-systems.com/

Documenting the main design decisions only is

the normal way to follow. By contrast, the ﬁrst on-

tology objective obeys to write a large documentation

where almost every concept or groups of concepts has

been justiﬁed.

3 ONTOLOGY DESIGN

The main concepts of the ontology are extracted and

justiﬁed from its own deﬁnition.

From a technical viewpoint, EIA is an analysis

process to identify cause-effect relationships and to

quantify, assess and prevent the environmental impact

of a project or an activity (G´omez, 2003).

As it is explained above, this structure will help

to environmental staff to understand the ontology or-

ganization in order to accomplish the ﬁrst objective,

using the ontology like reference knowledge. Hence,

the concepts should be grouped in a logic way.

Impact

PreventiveAction

Indicators &

MeasureUnits

ImpactAssessment ImpactedElement

InustrialActivities

ImpactingActions

ContaminantElement

Figure 1: Main concepts and relationships.

These important concepts are:

• Environmental impact which is any adverse or ad-

vantageous change in the environment produced

by the activities, products or services of an orga-

nization.

• Environmental elements which may suffer im-

pacts.

• Human actions which may produce impacts and

affect to the environment. These are the cause of

the impacts.

• Industrial activities, which are strongly related to

the human actions, although they have their own

troubles and a different abstraction level.

• Substances or contaminant elements. Although

these are not the cause or effect of an impact,

many times they are the way the impact is pro-

duced.

• Preventive actions, which are actions used to pre-

vent or hinder environmental damages.

KNOWLEDGE REPRESENTATION IN ENVIRONMENTAL IMPACT ASSESSMENT - A Case of Study with High

Level Requirements in Validation

413

• Environmental indicators or measure units for im-

pacts.

• Impact assessment, which are qualities linked to

the impacts.

These aforementioned concepts belong to the ﬁrst

hierarchical level although there are others concepts

and other levels below it.

The source used to extract the knowledge is not

so important for knowledge based systems although

it has more importance in the context of environ-

mental experts. These sources are books (G´omez,

2003; Garmendia et al., 2005; Smith, 2002; Block,

2000; Barettino et al., 2005); technical reports (Can-

ter and Sadler, ); other ontologies (GCMD, ); stan-

dards like UNE 150008 and UNE 14001; PhD Thesis

(Colomer, 2007; Garrido, 2008); Spanish legislation

like 1131/1988 and 11/2005; and international legis-

lation such as the European directives 2000/60/CE,

96/61/CE, 92/43/CEE, 79/409/CEE, 2455/2001/CE

and 85/337/CEE. In addition, a large range of bibli-

ography has been used but not referenced here.

3.1 Relationships and Concepts

A schematic diagram describing the most relevant

concepts and relationships among them is showed in

ﬁgure 1. These relationships are extracted instinc-

tively from the EIA deﬁnition as well as the concepts.

These are:

• hasPreventiveAction: An impact has a preventive

action.

• produceImpact: An industrial activity or an im-

pacting action produces an impact.

• hasIndicatorAndMeasureUnit: The indicator of

an impact or its measure unit.

• hasImpactAssessement: An impact has a speciﬁc

impact assessment.

• impactIn: An impact is produced over an environ-

mental factor.

The relationships allow the ontology improvement

adding knowledge and specifying formal deﬁnitions

of concepts.

These formal deﬁnitions are the base of any rea-

soning task and thus using the ontology to generate

new knowledge is possible. For example, the ontol-

ogy could be queried for the environmental indicators

of a series of impacts which are being evaluated.

The ontology is allocated for free access in the

web site http://arai.ugr.es/eiadifusa so that the whole

hierarchy can be explored.

4 WEB APPLICATION FOR

CONSULTING AND

CONTRIBUTIONS

An ontology with these characteristics needs to make

a strong effort in the evaluation phase. It has to in-

clude the concepts which really have interest in the

EIA process. This is the only way they could use

it like a knowledge base to give an assist with the

methodologies development by hand. Furthermore,

this is necessary to carry out the next step, to obtain

machine-driven development of EIA methodologies.

To sum up, the evaluation and the knowledge ho-

mogenization are the objectives of the web applica-

tion and this task should be done with the assistance

of environmental experts. However, they do not know

about OWL and requiring them to install applications

like Prot`eg`e tends to be problematic, because this is a

tool too much powerful for people who only have to

explore and check the ontology.

Moreover, a mechanism to collect the suggestions

should be made automatically to organize the infor-

mation properly and deal with a wide range of refer-

ees.

A web application

has been developed to collect

criticisms or suggestions and solve both problems.

The web interface is organized in two frames. The

ﬁrst one contains the concepts hierarchy and it allows

browsing it. The second one shows the concept infor-

mation when it is selected.

Two different kinds of contributions can be cho-

sen, related to a speciﬁc concept or with a general

character. Both ﬁll a simply form which contains in-

formation contact if it is necessary to establish com-

munication to clarify the criticism.

OWL

Web

Application

XML

Figure 2: Generation process of the web application.

The web application has been built according to ﬁgure

2 and using the own experience in the development of

a Final Application Generator based on model-driven

software development (Garrido et al., 2007).

First of all, the generation process begins pre-

processing the ontology in OWL. Then, an interme-

diate representation is produced in XML. And ﬁnally,

http://arai.ugr.es/eiadifusa

KEOD 2009 - International Conference on Knowledge Engineering and Ontology Development

414

all the ﬁles required by the php application are gener-

ated from the intermediate representation by a XSLT

transformation.

The routine work required to create a working

application is automated, thus avoiding the common

mistakes in a manual process; it would entail and im-

prove the overall application quality. Consequently,

if any change in the ontology or in the source ﬁles of

the web application is needed, it will be accomplished

and after that the web application will be generated

again.

5 FURTHER WORK

An ontology for EIA is built in order to provide ex-

perts with a reference for their methodologies devel-

opment. This use requires that it contains standard

and consensual knowledgeso that a web application is

development to allow the evaluation of their contents.

Hence, Environmental experts are called for contribu-

tions and suggestions, which will be later analyzed,

contacting directly with them or during conferences

related to EIA.

An application will be developed to give assis-

tance in the development process of methodologies.

The user will obtain the concepts related to the EIA

of a speciﬁc activity by querying the ontology.

Finally, as it mentioned before, the ontology will

be used to manage the knowledge of an EIA expert

which will provide information access, reasoning ca-

pabilities and a mechanism to assess the environmen-

tal impact of an industrial activity.

ACKNOWLEDGEMENTS

This research has been partly supported by projects

of ”Junta de Andaluc´ıa”: (CICE) P07-TIC-02913 and

P08-RNM-03584.

REFERENCES

Barettino, D., Loredo, J., and Pend´as, F., editors (2005).

Soil and water acidiﬁcation: problems and solutions

(in Spanish). Geological and mining Institute of

Spain.

Block, M. (2000). Identifying Environmental Aspects and

Impacts. ASQ Quality Press.

Brank, J., Grobelnik, M., and Mladeni, D. (2005). A sur-

vey of ontology evaluation techniques. In Proceedings

of the Conference on Data Mining and Data Ware-

houses.

Canter, L. and Sadler, B. A tool kit for effective EIA prac-

tice: review of methods and perspectives on their ap-

plication. a Supplementary Report of the International

Study of the Effectiveness of Environmental Assess-

ment. Technical report, University of Oklahoma, In-

ternational Association for Impact Assessment.

Colomer, F. (2007). Analysis and systematization of en-

vironmental security in urban landﬁlls. Application to

lixiviate ponds (in Spanish). PhD thesis, Valencia Uni-

versity.

Delgado, M., Duarte Velasco, O., and Requena, I. (2006).

An arithmetic approach for the computing with words

paradigm. International Journal of Intelligent Sys-

tems, 21(2):121–142.

Fern´andez, M., G´omez, A., Pazos, J., and Pazos, A. (1999).

Ontology of tasks and methods. IEEE Intelligent Sys-

tems and Their Applications, 14(1):37–46.

Gangemi, A., Catenacci, C., Ciaramita, M., and Lehmann,

J. (2006). Modelling ontology evaluation and valida-

tion. In ESWC, pages 140–154.

Garmendia, A., Salvador, A., Crespo, C., and Garmen-

dia, L. (2005). Environmental Impact Assessment (in

Spanish). Pearson Educaci´on.

Garrido, J., Martos, M. A., and Berzal, F. (2007). Model-

driven development using standard tools. In Proceed-

ings of the 9th international conference on enterprise

information systems (ICEIS 2007), pages 100 – 103.

Garrido, M. (2008). Environmental diagnostic methodology

in landﬁlls, adaptation to use with fuzzy thechniques

and to apply in Andaluc´ıa landﬁlls (in Spanish). PhD

thesis, Granada University.

GCMD. GCMD ontology, Semantic Web for

Earth and Environmental Terminology.

http://sweet.jpl.nasa.gov/ontology.

G´omez, A. and Benjamins, V. (1999). Overview of Knowl-

edge Sharing and Reuse Components: Ontologies and

Problem-Solving Methods. In proc. of the Nether-

lands/Belgium Conference on Artiﬁcial Intelligence

(BNAIC), page 242.

G´omez, D. (2003). Environmental Impact Assessment:

a preventive tool for environmental management (in

Spanish). Mundi-Prensa.

Schreiber, G., Wielinga, B., and Jansweiger, W. (1995). The

KACTUS view on the O word. In IJCAI Workshop on

Basic Ontological Issues in Knowledge Sharing.

Smith, K. (2002). Environmental Hazards, Assessing risk

and reducing disaster. Routledge.

Swartout, W., Patil, R., Russ, T., and Knight, K. (1997).

Toward distributed use of large-scale ontologies. In

Spring Symposium Series on Ontological Engineer-

ing, pages 33–44, Stanford. AAAI Press.

Uschold, M. and Gr¨uninger, M. (1996). Ontologies: prin-

ciples, methods, and applications. Knowledge Engi-

neering Review, 11(2):93–155.

KNOWLEDGE REPRESENTATION IN ENVIRONMENTAL IMPACT ASSESSMENT - A Case of Study with High

Level Requirements in Validation

415