Journal of Knowledge Management Practice, Vol. 11, No. 1, March 2010

Mapping Knowledge Into A Database Model

Emile Segev, Université du Québec, Canada

ABSTRACT:

In a highly competitive business environment, organizations are looking for improved tools which could provide them a greater opportunity to succeed and to create a strategic advantage in their market. Their primary concern therefore is continuous, effective and secure access to their accumulated knowledge. Knowledge Management tools and methods are emerging, primarily for the use of big organizations, but more and more small and medium enterprise (SME) are interested in adopting them. Knowledge is a step ahead of Information, and deals with the capturing and the gathering of Information along its steps and rules related to a Working Process, in order to perform the latter at an optimal level. The dilemma is, that in order to access knowledge, one must store it, in the most effective and efficient way.  This paper offers various definitions of Knowledge by differing authors, examines the general difficulties and the problems of Capturing and Storing Knowledge, and presents our particular attempt to map and store it into a database structure. It relates to the author’s current ongoing project, which objective is to convert a plain text knowledge system, into a common database system, for the use of a Property Management enterprise.  The motivation behind our attempt is, once knowledge is stored in such a structure, we believe it will attain our dual objective: an effective and efficient access to knowledge that is stored in a secure manner.

Keywords: Knowledge, Knowledge storage, Knowledge management, Database structure, DBMS


Importance Of Knowledge

We are in an era of Knowledge revolution, where knowledge occupies the center stage. Its continuous generation, sharing and implementation have become crucial for firms and countries. This revolution is supported by the revolution in information technologies (Pillania, 2008).  The latter is to support existing KM efforts in its various activities and possibly, to suggest newer and more effective ways to deploy them.  The relevance of KM is not argued by the academic or the practionners communities, nor is it a subject of demonstrations.  In today’s highly competitive business environment, organizations are looking for KM tools, which could provide them the highest effectiveness, a greater opportunity to succeed and to create a strategic advantage in their market.

What Is Knowledge?

To make sure we are in the right direction and using a common vocabulary, let’s examine some of the definitions of Knowledge.  In recent years, the term Knowledge Management has been used, to describe the efforts of organizations to capture, store, and deploy Knowledge (Preece et al, 2001). For some, Knowledge is a somewhat elusive concept; here is a pragmatic description of knowledge in organizations: Knowledge is a fluid mix of framed experience, values, contextual information, and expert insight that provides a framework for evaluating and incorporating new experiences and information. It originates and is applied in the minds of knowers. In organizations, it often becomes embedded not only in documents or repositories but also in organizational routines, processes, practices, and norms (Davenport and Prusak, 1998; 5).  We usually distinguish 'knowledge' from 'information', and information from 'data', on the basis of value-adding processes which transform raw material (for example, transaction records) into communicable messages (such as documents) and then into knowledge and other higher-order concepts. These value-adding processes include in the first instance contextualization, categorization, calculation, conversion and condensation; and in the second, connection, comparison, and conversation. Other authors - notably Thomas Stewart - dismiss the notion of a data-to-wisdom hierarchy as bogus and unhelpful on the grounds that "one man's knowledge is another man's data".(Stewart 1997; 69)

Knowledge Management And Its Facets

Another theme discussed amongst the experts is the distinction between the explicit Knowledge and the tacit knowledge. Here are some insights.  A more important distinction - which is fundamental to the concept of knowledge management - is that between 'explicit' and 'tacit' knowledge, explained by Ikujiro Nonaka.  "Explicit knowledge is formal and systematic. For this reason it can be easily communicated and shared, in product specifications or a scientific formula or a computer program. Tacit knowledge is highly personal. It is hard to formalize and therefore difficult, if not impossible, to communicate." (Nonaka, 1991)  

Most authors identify the different facets of KM as being one of Capturing, Organizing and Storing, while everyone clearly insists on the Sharing Aspect of KM.  It is obviously the goal of any KM efforts.

Ramjani (Ramjani, 2000) identifies the following steps or stages in KM:

Ø      Develop Knowledge (acquire, capture, create,,)

Ø      Preserve Knowledge (store, securing,,)

Ø      Update knowledge (evolving, improving,,)

Ø      Transfer knowledge (communication, deploying,,)

Ø      Transform knowledge (compiling, standardizing,,)

Ø      Assess knowledge (appraising, evaluating,,)

Ø      Apply knowledge (using, enacting,,)

The Case Of PMGT Inc.

In our attempt to map Knowledge into a Database Model, we used the knowledge of an actual organization located in Montreal, Canada.  PMGT inc. is a company that manages a real estate portfolio of about 1200 residential units, mostly rented to individuals.  The company has over the years gathered extensive knowledge and vast experience in the various processes evolved around the management of rental units. This involves several uncontrolled factors, such as the credit checks of the candidates, references from employers and previous landlords, and so forth. Another important external factor is the Government legislation involved with the leases signed by parties.  The lease itself, is a legal document and must obey to certain rules which are out of control, at the organization level. All these factors combined with the internal factors upon which the organization relies, make the Rental Management macro process, quite complex. 

From the beginning PMGT had to divide the entire process into sub-processes which are outlined later on this article.  All are presently well documented in various computer files, mostly Excel and plain text documents containing company’s procedures to follow.  This is the ‘’Knowledge Center’’ and anyone acting on behalf the company, must follow the instructions, and the prescriptions outlined in details in those files.  The Company also uses a standard Program for Property Management, used by many other Property management firms. The program is catered to respond to the country’s juridical environment and context, managed by a government body called the Rental Board.  These combined Resources are presently, the KM system of the company.  PMGT Inc. has been using its KM rudimentary system, for any given situation which must be addressed, according to the company policies.  The directors are very strict about following all the steps and actions, exactly as documented in the Procedures.  Along the course of its use, it was evident that the system had many flaws. First, the employees complained it was often difficult to locate the right Knowledge pertaining to a given situation.  At other times, it took a long time to find and access, and  in some cases, it was reported that the Knowledge used was incorrect and even erroneous.  This may be explained by the fact that the Knowledge stored was imprecise, incomplete, and very often, was not up-to-date.  These factors had negative repercussions on the quality of the decisions taken by the employees, to the point that it affected the global performance of the company in managing its properties in an effective way. 

In other words, Management did not have a method of ensuring accountability for actions and decisions taken by employees, since the actual system could not provide the authors and origin of the Knowledge, nor the subjects who enacted the actions and decisions.  At this point, the managers wanted to improve and bring their actual KM system, to  the Next level, as they called it. This is when we stepped in, to use this opportunity as an experimental base for our study.  We knew from the beginning that the resources of this small organization are not sufficient to maintain a very sophisticated KM program such as LOOM. 

PMGT Inc, has managed real estate residential properties for more than 25 years. Along the years, it has accumulated most valuable experience in the various processes involved in their activities.  The change and mobility of the  employees and the will to retain all their experience in place, has motivated the company  to invest in a Knowledge System where the objective is to capture the knowledge and store it as quickly as possible, in order to make it easily available to all employees.   PMGT has established an edge over its competitors in terms of effectiveness and it is now the time to build  a more structured KM tool .   Our goal is to bring forward the present rudimentary system  into an automated one which could be accessible in  effective and secure manners.  We must design and develop a more structured system, which will be mainly stored and supported by a Relational DBMS and later, if necessary, coupled with XML files.  Our priority is the storage aspect of all the actual knowledge, held in these documents.

Knowledge Storage

Most knowledge management activities are a combination of business processes and information technology (Bukowitz, W. and Williams, 1999).  In our KM system we are indeed using Business Processes, or as we call them, working processes. We investigated the literature and found that several authors have mentioned DBMS as means of Knowledge storage.  In terms of technology, most current knowledge management activities rely on database and internet systems: it is typically stored in databases either as simple tables or semi-structured text (for example, in Lotus Notes) (Preece et al, 2001).  In order to retrieve Knowledge in the best and most secure way, we must capture it in a system which allows rapid selective queries in a controlled environment. Therefore a DBMS is a good effective tool. Knowledge is broken down into permanent atomic "facts" which can be stored in a standard relational database and processed very efficiently. It also provides for the efficient querying of a knowledge base, efficient inference of new knowledge and translation into and out of natural language. Queries can also be processed with full natural language explanations of where the answers came from (Tunstall-Pedoe, 2006).

How Is KM Stored?

Knowledge, in all of its forms, whether explicit or tacit, includes rules, steps, actions and so forth, must be stored in some manner.

Ø      This order seems to be from the least to the most sophisticated tool.

Ø      Plain or structured (ex groupware program) text format

Ø      Hyperlink form using XML, used mainly within the Internet environment, XML is specialized to manipulate libraries.  XML's purpose is to aid information systems in sharing structured data, especially via the Internet, (Bray et Al, 2006).  The data is defined in a Hierarchal format, which enables, for example, a library, to input the reference data of all its the books. The user interface for entering the data is in a form of a Table. XML,  the interface,  will be a table, but the storage is Internet compatible using hyperlink, to access another structure, related to the present information linked.. 

Ø      Database structures

Ø      Sophisticated and costly KM representation systems such as Classic, Loom or G2, which are aimed for the larger scale organizations. 

Ø      Combination of the above.

Choice For KM Storage

From the start, we knew that it was an experimental project and as a SME, the option of the more sophisticated options, were discarded. There is little use of sophisticated knowledge representation systems such as Classic, Loom, or G2. Few organizations have a systematic process for capturing knowledge, as distinct from their conventional information-capture procedures (Preece et al 2001).  Based on this premise, and based on the author’s experience, a lighter solution was wanted, an inner solution, a self sufficient one, so as not to depend on any outside resources, using tools, unfamiliar to the employees.  We also decided that a solution based on XML platform was not necessary at the moment.   We had some evidence that KM can be stored and managed by a DBMS.  However, we did not exclude the possibility, that in the course of the project, we may have to combine our system to an XML solution and build a more intelligent, effective and efficient tools, embodying the security of the very confidentiality of the Knowledge, captured in the system.  Hence our decision was to start with a simple prototype involving one Process only. We chose the most important one for PMGT that is the Application Rental Apartment, (codified ARA).

Methodology

Ø      We have the list of all working processes, that is:  P(A,1) to P(A,n)

Ø      Our Methodology is cautious and will follow this procedure:

Ø      Keep and  Maintain the present system in use

Ø      Develop  P(A,1) to  P(M,1)

Ø      Test P(M,1) in an operational working environment

Ø      Bring the necessary corrections and modifications and validate P(M,1)

Ø      Replace P(A,1)  by P(M,1) 

Ø      Iterate  procedure to all  remaining  P(A, n)

Ø      Once the cycle is finished, the new system would replace the existing one and would become the operational KM System in use by PMGT.

How Did We Proceed?

At first, we had to define what are the facts, steps, rules, actions, documented and undocumented information which were relevant to the Process in study. We had to gather all the tacit knowledge which is ‘’nowhere’’ seen i.e. it is everything that is said among all conversations of the employees, not only when officially working but also when they exchange impressions and experiences of work habits, according to context.  For example, in the process that we are in the course of developing ARA,  there are these facts which we captured from the employees, as being tacit knowledge, which is in many ways, intuitive.  It  is complex and multifaceted process as it may involve the engaging of various types of juridical procedures. The employee, in charge of this process says: “It may be a very short or a very long, time consuming and complicated process”, as we will see later.

Capturing And Retaining PMGT Knowledge

In this exercise, we must first ‘’extract ‘’ the tacit and non tacit knowledge from the individuals who retain it to be captured in a more structured manner. Focus would be on the employees who possess the most experience and who have been acting to perform the process on behalf of the organization. They often have intuitive   knowledge based on their experience performing the same Process, under different conditions, situations and contexts. Their valuable experience must be expressed, captured, classified and stored, into our data structure model.  We may, along the way, modify the Database structure in order to address tasks or relations, in the process, which are difficult or impossible to store in the current structure, and vice versa. We adapt the Process components to the DB structure, and we may at some situations adapt the structure by performing several alterations, in order to be able to integrate the components of the Process.

Implementing The KM System In A DB Structure

 

 

 

 

 


Figure 1: The Model Used To Translate And Store The Mental Process ¨Extracted¨ From An Experienced Worker

Our structured system is based on this model, whereas the Mental Process is translated into a structured form which in turn, is converted into Steps Rules and Actions, as illustrated in the above figure. This could be expressed as follows:

PM → PS:  {Steps, (Rules), Actions}

PM  and  PS, being, the Mental and Structured process respectively.

PMi  PSi :  { Step(PMi) + (Rules (PMi)) + Actions(PMi)}

Process i = (Step1 + Step2 +… Stepn  )

Stepi = (Rule1   + Rule2 + ………..Rulen  )  +  (Action1  + Action2 + …………..+ Actionn )

Hence,

Process i  = (Stepi……. Stepn ) + (Rulei ……..Rulen ) + (Actioni …………..+ Actionn )

 or

Processi = ∑ (Stepi to n,  (Rulei to n ), Actioni to n )

We will see later, that Rules are optional, as we may have situations consisting of a sequence of Actions, with no Rules attached.

The Data Base Content

The following DB model is a simplified content for the application being studied in this project.  First, we define the Processes involved in the application, then we present in a more detailed manner one of the Processes.  Our hypothesis is: once the system is validated throughout the said Process, it would most probably function on the others. The only difference between the Processes is their respective number of steps and their complexity, as expressed by the set of Rules involved. For the purpose of this article we present only ARA- Step-5.

T_Process

Code-P

Description

ARA

Application Rental Apartment

ARC

Application Rental Commercial Unit

RCL

Rent Collection

TLR

Tenant Late paying the Rent

CMT

Tenant Call For Maintenance

TSK

Tenant Skipping

TDM

Tenant  causes Damages to Unit

TLV

Tenant wants to leave & break Lease

ARI

Annual Rental Increase

 

T_Process_Steps

 

Code_P

Step

Description

Code_F

ARA

1

Applicant fills Form and Sign

FRM1

ARA

2

Check all fields and Instructions on the form

 

ARA

3

Enter Form in RMS

 

ARA

4

Send Form-RMS by Internet to credit Check

FRM1

ARA

5

Reception &process Credit Check response

 

ARA

6

Fill & Sign LA

FRM4

 

T_Forms

 

Code-F

Description

Nb Pages

FRM1

Printed form Applicant for residential unit fills

1

FRM2

Notice of Lease Renewal

2

FRM3

Application to End the Lease

3

FRM4

Lease Application

6

 

T_Rules_Set

 

Code_P

Step

Code_R

Description

Value

Units

Other_1

A-1

A-2

A-3

A-4

ARA

5

R1

CC Negative

 

 

 

A1

A2

 

 

ARA

5

R2

Lease  Less

12

MM

 

A3

A4

A5

A6

ARA

5

R3

Lease  GE

12

MM

 

A3

A4

A5

A7

ARA

5

 

 

 

 

 

A8

 

 

 

 

 

 

 

 

 

 

 

 

 

 

CC=Credit Check

 

T_Actions

 

Code-A

Description

 Approval

A1

Reject Application

Agent

A2

Send response to Client

Peer

A3

Print  Lease

Peer

A4

Applicant signs

Peer

A6

PMT 3 first months rent

Accountant

A7

PMT 1 month rent

Accountant

A8

Approval

Superior

We present here a simplified structure, which could be used as a base in designing a small KM system in a Relational Model.   Fields could be defined as Hyperlink to a Document. The latter may be text, audio, video and a mixture of these media.  We present a scheme which could summarize the structure of the proposed Data Base.

 

 

 

 

 

 

 

 

 

 

 

 

 

 


Figure 2:  The Global Scheme Including The Cardinalities, Of The Database Model

The Global scheme reflects and validates the illustrated model as presented in figure-1. We can see the Process Steps being either translated directly into Actions or into Rules, in order to be transformed and broken down into Actions.  The Rules in this case, are, the Process Filters, before they are broken down into Actions.  This Scheme is not in a Normalized form.  Our intention was to make it as clear as possible for the reader. The Normalized scheme for this Data structure could be the following:

 

 

 

 

 

 

 

 

 

 

 

 


Figure 3:  The Global Scheme And Cardinalities, Of The Database Model, In A Normalized Form

Conclusion

In Today’s business world, vast amounts of data and information are filtered through an organization. A method for representing, capturing, storing and retrieving real-world knowledge on a DBMS structure is presented here.  We described in detail the model and its content, in a simple manner.  In this experimental study, the goal of the author is not the relevance of the various KM tools, but rather to suggest a simple and rapid solution to a small organization, in need of a more structured KM.  DBMS are common tools and are easily and rapidly implemented.  The small organization lacks technical resources and could not manage sophisticated and cumbersome systems, catered exclusively for the Management of Knowledge.   This could be viewed as a transit solution till the more sophisticated tools could be applied. The model, need to be presented in a more detail form before the Implementation phase, in which, we intend to put these concepts into a RDBMS. It could be the subject of a subsequent article. We will present, in a more comprehensive way the Design of the data structure, as well as some forms of complementary structures, such as XML or others.

References

Bray, T. Paoli, J., Sperberg-McQueen, C.M, Maler, E. and Yergeau, F. (September 2006). "Extensible Markup Language (XML) 1.0 (Fourth Edition)-Origin and Goals". World Wide Web Consortium; retrieved on October 29, 2006.

Bukowitz, W. and Williams, R. (1999) Knowledge Management Fieldbook,Prentice-Hall.

Davenport, T. H. and Prusak, L. (1998) Working knowledge: how organizations manage what they know. Boston, Ma: Harvard Business School Press.

Harvard Business Review on Knowledge Management, Harvard Business School Press, 1998.

Nonaka, I. (1991) The knowledge creating company. Harvard Business Review, 69 (6), 96-104

Preece, A., Flett, A., Sleeman, D., Curry, D., Meany, N. and Perry, P. (2001) Better Knowledge Management Through Knowledge Engineering: A Case Study in Drilling Optimisation, Aberdeen, UK.

Pillania, R.K. (2006) State-of-Art of Knowledge Storage and Access in Indian Industry (July 13, 2008). Journal of Information & Knowledge Management, 5 (1), pp. 55-61; Available at SSRN: http://ssrn.com/abstract=1159386

Rajamani, U. (2000). Unpublished material

Stewart, T. A. (1997) Intellectual capital: the new wealth of organizations. London: Nicholas Brealey.

Tunstall-Pedoe, W. (Cambridge, GB) (2006) Knowledge storage and retrieval system and method, United States Patent 7013308 , Inventors: Application Number:09/990188; Publication Date:03/14/2006


About the Author:

Dr. Emile Segev has instructed as an Adjunct Professor at the University of Québec of Montreal and at Ecole de Technologie Superieur. He has taught Computer Science and MIS courses in the department of Management and Technology.  Dr. Segev has written three books in the field of Databases. The most recent are: "Les Base de Données: C'est quoi et Comment les concevoir": (Database Systems: What are they, and How to Design Them) and “Les Bases de Données: les concevoir et les réaliser" (Designing and Developing Database Systems). Dr Segev is a practitioner in the field of Computer Systems and he leads web-based transactional projects for www.webus.ca. During the course of his consulting assignments, he has led several computer projects abroad, mainly in South America, China and Vietnam.