Using Ontology To Define Decision Problems Of Enterprises

Using Ontology To Define Decision Problems Of Enterprises

Anass Mamouny1, Fatima-Zzahra Cheffap, Mostafa Hanoune3 Doctoral Studies: Mathematics, Computer Science and Information Technologies Hassan II Mohammedia-Casablanca University, Faculty of Sciences Ben M'sik

BP 7955 Casablanca, Morocco

Abstract

Competitive Intelligence (CI) allows companies to collect information relevantly and to treat them while ensuring the legality and safety. The definition phase of information needs is an essential step in the process of CI, as it identifies the field to monitor the business and define the scope of the strategic watch. We propose to make a model for the business and its internal and external environment to help decision makers and watchers in the definition of decision problems. We choose the notion of ontology as a way of modeling, as it provides the tools to describe explicitly the conceptualization of knowledge in a specific field. Our goal is to build enterprise ontology which can be used in companies for the definition of information needs.

1. Introduction

   To cope with the constant change of economic, geopolitical and technological environment, company should be able to adapt to changes caused by demands of its customers, attitudes of its partners and behaviors of its competitors. Competitive Intelligence (CI) is the most effective tool for the company to find useful information, and to develop the strategy for achieving its objectives. CI is a continuous cycle, whose steps include: (a) Identify information needs and decision problems across the firm, (b) Collect information (conducted legally and ethically), (c) Analyze information and upgrade it to intelligence, (d) Disseminate results to decision-makers, (e) Identify feedback of decision makers and their new needs.

   Analysis of CI process shows that techniques and methods used in this approach are useless if the results are incompatible with the needs and strategic objectives of the company. Therefore, the identification of informational needs must be conducted with great interest. Many models and methods (i.e. SWOT, PORTER) are used in enterprises to define the

   decision-making needs but they do not cover the overall scope of the company and must be combined to give to firms the global strategic analysis. The purpose of this paper is to build an enterprise ontology named EntrOnto for modeling the internal and external environment of the firms, and to show the role of this ontology in Competitive Intelligence process.

   According to Fernandez and al. (1997), the construction step moves into following states: specification, conceptualization, formalization, implementation, and maintenance. We followed the same steps in construction of EntrOnto. The first part of our paper will concentrate on knowledge domain and terminology of our ontology. In the second part, we present the work done on development of EntrOnto. Finally, we demonstrate how to use our model in definition phase of the decision problem of companies.

2. Knowledge Domain of EntrOnto

   In this section we describe the various parts of the Knowledge Domain of EntrOnto: Internal-Environment and External-Environment of firms. We divided the External-Environment into two parts: MicroEnvironment and MacroEnvironment.

   1. Internal-Environment

      To describe the internal environment of an enterprise, we must describe internal factors and common variables to all firms i.e. hierarchy, structure, human resources, means of communication, goals, etc. Terminology and structure (concepts and relationships) of the organization-ontology developed in "TOVE" project meets this description. In this ontology, an organization is a set of constraints on activities performed by agents. Each agent plays one or more roles, and he is a member of a division or a team. He consumes a range of resources (e.g. materials, labors, tools, etc.) to perform its activities, and he has a set of communication-link dening the protocol that it communicates with other agents in the organization.

      Moreover, each role has a set of organization-goals and it requires a set of skills and authorities to achieve the goals. After our analysis, we deduced that organization-ontology includes most of internal factors which are useful for decision-makers to define decision problems of their firms. We decided to integrate the organization-ontology as it is in our ontology for its accuracy. In the first paragraph of the second section, we will present its structure (concepts and relationships) and how we linked it with the other parts of EntrOnto.

   2. MicroEnvironment

      MicroEnvironment part describes the variables of immediate environment of firms. For Michael Porter, the competition is not limited to only direct competitors. To maintain profits, the firm will have to face four competitive strengths:

      • New Entrants: They are looking to quickly make their market position.

      • Customers: They have power to negotiate the purchase price, conditions and terms of payment and the quality of service and delivery.

      • Suppliers: They have power of pressure, including price, quality and quantity of services.

      • Substitute Products or Services: They lower costs without compromising quality or impose a high cost of replacement.

        Our ontology includes PORTER forces in his conception because they are useful in step of

        requirements definition in CI process. We chose the terms: customer, supplier, competitor, partner and product for modeling Porter forces, and the terms: has_custumer, has_supplier, has_competitor, has_partner and buy for modeling their relationships with the company.

   3. MacroEnvironment

      MacroEnvironment part contains global variables that affect the company. PESTEL analysis draws the outlines of the enterprise environment by six key criteria: Political, Economic, Socio-cultural, Technological, Ecological and Legal. This type of analysis provides the ability to objectively display the variables that may affect the strategic operations of enterprises. We chose the term macroenvironment

      _factor for class that represent all PESTEL factors and the term macroenvironment _category for class that represent all possible categories of macroenvironment

      _factors. Then, we chose the term affects for the relationship that connects macroenvironment_factor to organization. Decision makers will use this relation to specify the influence type and the influence degree of macroenvironment _factor. And we chose the term belongs for relationship that connects macroenvironment_factor to macroenvironment

      _category. It will be used for classifying the factors in their own categories.

      Figure 1. Diagram of binary relations created on conceptualization step of EntrOnto Ontology.

3. Building EntrOnto

   3.1 Conceptualization Step

   We summarize the work done in this step in a diagram of binary relations (Figure 1). In this diagram, the concepts are represented by rectangles and the relationships by arcs (from source concept to target concept). The choice of these concepts and these relationships is taken from the concepts and the models that we presented in the first part. With this structure, we guarantee to decision makers an overview of all the elements that can help them to find decision problems that they should solve to improve the situation of their companies.

   Each concept and each relation of EntrOnto has a set of attributes. We give some examples in the following table:

   Table 1. ome Attributes of EntrOnto

   ATTRIBUTE	CONCEPT	TYPE
   Short_Name	Organization-agent	Character string
   Type	Organization	Character string
   Country	Organization,	Character string
   Creation_Date	Product,	Date
   Employees_	Organization,	Integer
   Number	Division
   Turnover	Organization,	Float
   Website	Division	Character string
   Address		Character string
   Social_	Organization	Enumeration
   Situation	Organization
   Telephone	Organization	Character string
   Experience	Organization-agent	Float
   Start_Date		Date
   End_Date	Organization-agent	Date
   Price	Organization-agent	Float
   	Activity
   	Activity
   	Product

   1. Formalization Step

      There are currently several formalisms for representing ontologies, such as conceptual graphs (CG) and description logics. A conceptual graph is a graph composed of two types of nodes: concepts and conceptual relations, and arcs are used for linking. In graphical form of CG, concepts and conceptual relations are represented by rectangles, and relationships with circles and arcs are used to link the relationship with concepts associated. The graphic form of the CG makes it easy to interpret, modify or create new ones. This facility helped us in our ontology. The following example shows an extract of our formalization which shows the relation between organization, customer and product.

      Figure 2. Extract from Conceptual Graph created on

      Formalization step of EntrOnto, concepts are: organization, product and customer, and conceptual- relations are: produce and buy.

   2. Encoding Step

   For coding, the ontology must be formulated with a knowledge representation language such as OWL. OWL (Web Ontology Language) is used to describe the ontologies by setting the classes, subclasses, properties and sub-properties, relations and instances. We use Protégé-4 Framework for creation and editing of our ontology. Protégé is a system developed in Java by collaboration between Stanford University and Manchester University. It is free and open source and provides a flexible user interface. The validation is done by Fact ++ reasoner integrated with Protégé-4 to infer new facts and to verify the consistency of our model. The following figures show the class hierarchy and relation hierarchy of our ontology on Protégé-4.

   Figure 3. Class Hierarchy and Relation Hierarchy of EntrOnto on Protégé-4

4. Using EntrOnto

   1. Watchers Tasks

      The Watchers are the managers of research, processing and analysis of information sources for responding to needs of decision makers. The Watchers are looking in various information sources and they communicate the results to decision makers in order to make the right decisions for improving the economic situation of their companies. EntrOnto provides to watchers a base of informational and strategic needs which are common between most of firms. The watchers are responsible for:

      • Installing Protégé-4 tool and implement EntrOnto to make it available for decision makers of their firms.

      • Creation instances/individuals of classes, relationships and attributes.

      • Adding classes or relationships on EntrOnto those appear useful.

      • Maintaining EntrOnto and check its consistency after each change.

      Individuals (instances) are the basic, "ground level" components of ontology. They include concrete objects and they give to decision makers a knowledge base of its internal and external environment.

   2. Decision Makers Tasks

      The role of decision maker in competitive intelligence process is to identify decision problem and to define informational needs for watchers and to analyze the results in order to find solutions. To perform this step, the decision maker must be a person who knows and who masters the environment of his firm. Entronto is the tool which provides the necessary knowledge base to decision makers in order to detect issues that can become decision problems. This tool highlights the various relations between the variables of enterprise environment which are not evident and which can give more opportunities to firms. Decision makers can access to EntrOnto using the "OntoGraph" tab of Protégé-4 tool which visualize graphically the concepts and their relationships. The Figure 4 is an export of extract of EntrOnto from "OntoGraph" tab of Protégé-4. Decision-makers can find and give the decision problems listed on Table 2 using the individuals illustrated on Figure 4.

      The figure (Figure 4) shows some instances of tow concepts of EntrOnto: macroenvironment_ factor and macroenvironment _category.

      Figure 4. Example of some instances of macroenvironment_factor class and macroenvironment _category class.

      Table 2. Some Decision problems interpreted from instances of EntrOnto

      Instance	Decision Problem deduced
      Demographics	What is the population's growth rate
      	and age profile in client countries?
      Media_Views	What are the popular media in
      	customer country can be used to
      	advertise our product?
      Tax_Policy	Could any pending legislation or
      	taxation changes affect our business,
      	either positively or negatively?
      Unemployement	What is the unemployment rate? Will
      	it be easy to build a skilled workforce?
      New_discoveries	Are there any new technologies on the
      	horizon that could radically affect our
      	business?
      Climate	How climate could changes the
      	attitude of our customers?

5. Conclusion

   In this paper, we suggest an ontology which models the enterprise domain and its environment. It will help managers in phase of identification of informational and decisional problem. This is a sub goal of our goal which is the construction of a dedicated CI platform for companies using the new techniques of Web Semantic. We hope eventually enhance our ontology by inference rules using SWRL language to improve the processing of automated reasoning. These rules will allow the systems to perform logical reasoning called inferences and reaching conclusions that can assist managers in its decisions. The next step is the conception of the rule base and the implementation of this base in EntrOnto.

6. References

1. M. Uschold and M. King. Towards a methodology for building ontologies. Workshop on Basic Ontological Issues in Knowledge Sharing, IJCAI-95, 1995.

2. K.B. Akinbode, O.B. Longe, and O.I. Omoifo. Constructing ontologies in owl using protégé 4. African Journal of Computing & ICT Reference, Vol. 4. No. 3:2326, 2011.

3. N. Bouaka and A. David. A proposal of a decision- maker problem to for a better understanding of information needs. 1st International Conference on Information & Communication Technologies: from theory to Applications, ICTTA04. (Damas, Syrie):5, 2004.

4. TOVE Ontology Project. Availaible: http://www. eil.utoronto.ca/enterprisemodelling/tove/

5. A. Haberberg and A. Rieple. Strategic Management: Theory and Application. Oxford University Press, 2007.

6. B.J. Jaworski, L.C. Wee, and Society of Competitive Intelligence Professionals. Competitive Intelligence: Creating Value for the Organization. Society of Competitive Intelligence Professionals, 1993.

7. N. Juristo, M. Fernandez-Lopez, and A. Gomez- Perez. Methontology: from ontological art towards ontological engineering. Spring symposium series on ontological engineering, AAAI-97, 1997.

8. M. Gruninger M.S. Fox, M. Barbuceanu and J. Lin. An organization ontology for enterprise modelling. Simulating Organizations: Computational Models of Institutions and Groups, M. Prietula, K. Carley & L. Gasser (Eds), Menlo Park CA: AAAI/MIT Press: 131 152, 1998.

9. H. Martre, P. Clerc, France. Groupe "Intelligence économique et stratégie des entreprises.", France. Commissariat général du plan, and C. Harbulot. Intelligence économique et stratégie des entreprises. 1994.

10. M.E. Porter. Competitive Strategy: Techniques for Analyzing Industries and Competitors. Free Press, 2008.

11. M.E Porter. How competitive forces shape strategy. Harvard business Review 57, no. 2:137145, March/April 1979.

12. Owl web ontology language. http://www

    .w3.org/TR/owl-features/

13. M. Fernandez-Lopez and O. Corcho. Ontological Engineering: with examples from the areas of Knowledge Management, e-Commerce and the Semantic Web. First Edition. Advanced Information and Knowledge Processing. Springer, 2010.

14. P. Buche and J. Dibie-Barthelemy and L. Ibanescu. Ontology Mapping Using Fuzzy Conceptual Graphs and Rules. Supplementary Proceedings of the 16th International Conference on Conceptual Structures, ICCS 2008, Toulouse, France, July 7-11, 2008.

15. H. Knublauch, M. OConnor, S. Tu and M. Musen, Writing Rules for the Semantic Web Using SWRL and Jess. 8th International Protege Conference, Protege with Rules Workshop, Madrid, Spain, 2005.