DATABASE MARKETING PROCESS SUPPORTED BY

ONTOLOGIES

System Architecture Proposal

Filipe Mota Pinto

, Alzira Marques

Departamento Engª informática - Instituto Politécnico Leiria

Morro do Lena – 2411-901 Leiria, Portugal

Departamento Gestão e Economia - Instituto Politécnico Leiria

Morro do Lena – 2411-901 Leiria, Portugal

Manuel Filipe Santos

Departamento Sistemas de Informação – Escola de Engenharia

Universidade do Minho, 4800-058 Guimarães Portugal

Keywords: Ontologies, Database Marketing, Knowledge Extraction Process, Action Research.

Abstract: This work proposes an ontology based system architecture which works as developer guide to a database

marketing practitioner. Actually marketing departments handles daily with a great volume of data which are

normally task or marketing activity dependent. This sometimes requires specific knowledge background and

framework. This article aims to introduce an unexplored research at Database Marketing: the ontological

approach to the Database Marketing process. Here we propose a generic framework supported by ontologies

and knowledge extraction from databases techniques. Therefore this paper has two purposes: to integrate

ontological approach in Database Marketing and to create domain ontology with a knowledge base that will

enhance the entire process at both levels: marketing and knowledge extraction techniques. Our work is

based in the Action Research methodology. At the end of this research we present some experiments in

order to illustrate how knowledge base works and how can it be useful to user.

1 INTRODUCTION

Database Marketing (DBM) is a database oriented

process that explores database information in order

to support marketing activities and/or decisions. The

Knowledge Discovery in Databases (KDD) process

is well established among scientific community as a

three phase process: data preparation, data mining

and deployment/evaluation. The KDD has been

successfully applied in various domains particularly

in the marketing field. Nevertheless previous well

established concepts and scientific dominance

regarding each one of these methods, seem to have a

lack of knowledge concerning its application

amongst different requirements and conditions.

Available literature describe a DBM project as

comprised of a sequence of phases and highlight the

particular tasks and their corresponding activities to

be performed during each of the phases. It seems

that the large number of tasks and activities, often

presented in a checklist manner, are cumbersome to

implement and may explain why all the

recommended tasks are not always formally

implemented. Additionally, there is often little

guidance provided towards how to implement a

particular task These issues seem to be especially

dominant in case of more complex analytical

objectives at marketing activity understanding phase

which is the foundational phase of any DBM project.

In computer science, ontologies provide a shared

understanding of knowledge about a particular

domain (Gruber, 1993). At the best of our

knowledge the number of contributions to the

construction of marketing ontologies is very low.

However, they are starting to come to light through

407

Mota Pinto F., Marques A. and Santos M. (2009).

DATABASE MARKETING PROCESS SUPPORTED BY ONTOLOGIES - System Architecture Proposal.

In Proceedings of the 11th International Conference on Enterprise Information Systems - Information Systems Analysis and Speciﬁcation, pages

407-410

DOI: 10.5220/0002011704070410

 SciTePress

some marketing or computer research centers

(Grassl, 1999), (Bouquet et al., 2002), (Zhou et al.,

2006).

This research is part of a larger project to build

and develop a DBM Ontology (DBMO). The

DBMO should cover a semantic description of

processes supporting DBM, comprising classified

marketing objectives and activities, knowledge

extractions methods, objectives and tasks.

Our proposed research context focuses DBM as

the intersection of two others disciplines (knowledge

extraction techniques and marketing). Here, we

introduce ontologies as support to the knowledge

structure and integration of both.

One of the promising interests of marketing

ontologies is their use for guiding the process of

knowledge extraction in DBM projects. A tool that

gradually accumulates knowledge of the previous

domain developed processes is appropriate due its

iterative nature. Researchers often rework their data

in order to optimize further interactions. Integrating

this knowledge with ontology extends the ontology

usefulness.

We are proposing the initial conceptual structure

to the domain ontology as an integral part of a global

marketing system. According to some researchers

our ontology can be classified as an application

ontology (Sowa, 2000), serving our main global

project.

2 THE USE OF ONTOLOGIES IN

MARKETING

Ontologies are nowadays one of the most popular

knowledge representation techniques. When

ontologies are formalized in any kind of logic

representation, they can also support inference

mechanisms (Mylopoulos et al., 2004). For a given

collection of facts, these mechanisms can be used to

derive new facts or check for consistency. Such

computational aids are clearly useful for knowledge

management, especially when dealing with complex

and heterogeneous knowledge problems or with

large amounts of knowledge.

Ontologies use a formal domain or knowledge

representation, agreed by consensus and shared by

an entire community. Ontologies roles in DBM have

particular significance in a cross research (both

marketing and extraction techniques knowledge is

needed) area focus. Indeed, ontologies can play an

important role describing in a semantic form, all

concepts and techniques around the process.

Moreover, with such description it will also be

possible, to introduce metrics to compare and

therefore select and suggest the best approaches and

methods to a new project.

3 RESEARCH APPROACH

We have used action research based on two main

reasons. Firstly, due the low number of scientific

research works that has been conducted on

supporting DBM process over intelligent structures

like ontologies, the process by which this may be

completed is unclear. Secondly, ontologies can play

an important role in the knowledge development as

long as they register past knowledge for future reuse.

Thus exploratory research was required and action

research provides this capability better than many

other alternatives (Dick, 2008). Action Research

approach develops in a four step framework: first

formulate (plan), test (act), deploy (observe) and

evaluate (reflect). In this work we introduce a

connection element between each interaction:

ontological support. Supported by a previous

research work throughout the marketing knowledge

we had constructed a symbolic model for

representing knowledge and a tree structure (Figure

1). Here we intended to differentiate between

different knowledge levels structure tree towards the

following statements: Principal data information

type identification in marketing database; Main

DBM steps from marketing data to customer

knowledge; and DBM process’ matrix: Knowledge

base elements identification and creation

4 FINDINGS

The research project was done with a group of

database marketing practitioners. Our preliminary

findings are summarized in Table 1. Ending the

action research a practical and functional analysis

was made towards a possible conceptual semantic

map. Turning our action research to analytic

generalization, we can build a theoretical framework

(Yin, 2003). Linked to extant literature that shows

how the DBM process is developed, how associated

marketing knowledge can be structured and which

knowledge discovery approaches may be used.

ICEIS 2009 - International Conference on Enterprise Information Systems

408

Thing

Data

Extraction

Process

Relationship

Marketing

Activities

Type

Objectives

Data

Selection

Data

Pre-processing

Modelling

Evaluation

Source

Value

Market

Trigger

Personal

Market

Financial

Social

Life

Other

consumer

Transactional

LifeStyle

Demographic

Psychographic

String

Internal

External

Logical

Number

Date

Objective

Pre-processing

Algorithm

Customize

Interact

Diferentiate

Identify

Retention

Fidelization

Churn

is-a

Figure 1: Current Database Marketing Ontology.

Table 1: Action research findings.

Research issue Findings about the research issues

Principal marketing

database data type

information

From some literature review and supported by previous work done we

found four main marketing data types: Psychographics; Demographics;

Life style; and, Transactional.

Main DBM steps Based on both practice and literature review we considered the following

steps as a stable DBM process framework: Marketing objectives

definition and activity selection; Data selection; Data preparation; Data

pre-processing; Modelling; Model evaluation; and Business deployment

and evaluation.

DBM process’ matrix

Marketing

Objectives | Activities

Knowledge Extraction

cases

Description record set:

{

Data set

Data selection

Data pre-processing

Data Preparation

Algorithm used

Technical evaluation

Business evaluation

}

DATABASE MARKETING PROCESS SUPPORTED BY ONTOLOGIES - System Architecture Proposal

409

5 DISCUSSION AND

CONCLUSIONS

The extent, degree and simplicity of communication

enabled by the ontology makes it a synergistic

component of DBM strategy. An ontological DBM

approach solution appears promising for both

marketers and computer scientists.

One of the promising interests of DBM

ontologies is its use for guiding the process of

knowledge extraction from marketing databases.

This idea seems to be much more realistic now that

semantic web advances have given rise to common

standards and technologies for expressing and

sharing ontologies (Coulet et al., 2008). In this way

DBM can take advantage of domain knowledge

embedded in DBMO. The results of this research

have implications for both theory and practice. The

first practical results relate the possible feedback

between different DBM projects through a table with

all used resources registered. It will be possible to

implement, through ontologies, a knowledge base

with suggestion or work profile capability. That

Knowledge base, according to the previous

registered experiments will be also capable to

suggest to each marketing objective which

marketing activities, data to be selected and also

tasks to be performed should be chosen. Another

implication relates to the benefits of a global view of

marketing databases role in marketing objectives:

then is possible to fill them with appropriate data.

Our model further emphasizes the importance of

the marketing knowledge to be structured in order to

allow resources reuse or even to achieve synergies in

marketing activities development. Thus managers

and marketers should be aware of this issue, because

there is a loop through which performance of DBM

process can effectively be improved.

The research findings and contributions have

several implications for the theory about ontologies

and DBM, as well as the use of Action Research

methodology. This research provides new insights

into DBM theory in two ways: First this research

appears to provide the first global investigation

about the intersection of ontologies and DBM in

organizations, and how it may be achieved. Thus

this research contributed to the theory-deficient area

of the integration of ontologies and DBM. Second

there is to few literature dedicated to marketing

ontologies and thus this research appears to be one

of the first academic investigation of this

phenomenon.

The impact of that ontology is the future

initiation to a shared DBM knowledge platform that

will provide a trusted base between marketers, DBM

practitioners and artificial intelligence researchers.

Indeed this research identifies a number of areas

requiring further research, namely to marketing

knowledge tree and therefore marketing ontology.

REFERENCES

Bouquet, P., Dona, A., Serafini, L., and Zanobini, S.

(2002). Contextualized local ontology specification via

ctxml. In for Artificial Intelligence, A. A., editor,

MeaN-02 AAAI Workshop on Meaning Negotiation,

Edmonton, Alberta, Canada. AAAI.

Coulet, A., Smail-Tabbone, M., Benlian, P., Napoli, A.,

and Devignes, M.-D. (2008). Ontology-guided data

preparation for discovering genotype-phenotype

relationships. Journal of BMC Bioinformatics, 9:1–9.

Dick, B. (2008). Postgraduate programs using action

research in action learning, action research and process

management: Theory, practice, praxis. Technical

report, Faculty of Education, Griffith University.

Grassl, W. (1999). The reality of brands: Towards an

ontology of marketing. American Journal of

Economics and Sociology, 58(2):313–319.

Gruber, T. R. (1993). A translation approach to portable

ontology specifications. Knowledge Acquisition,

5:199–220.

Mylopoulos, J., Jurisica, I., and Yu, E. (2004). Ontologies

for knowledge management. Knowledge and

Information Systems, 3:380–401.

Sowa, J. F. (2000). Knowledge Representation: Logical,

Philosophical, and Computational Foundations.

Brooks Cole Publishing Co.

Yin, R. K. (2003). Applications of case study research,

volume 34 of Applied social research methods series.

Sage Publications, Thousand Oaks, CA.

Zhou, X., Geller, J., and Halper, Y. P. M. (2006). An

Application Intersection Marketing Ontology, chapter

Theoretical Computer Science, pages 143–163.

Lecture Notes in Computer Science. Springer Berlin /

Heidelberg.

ICEIS 2009 - International Conference on Enterprise Information Systems

410