404  

An Integrated Framework for Knowledge Audit and Capture

Mohd Sharifuddin Ahmad, Mohd Zaliman Mohd Yusof, Azhana Ahmad

College of Information Technology,

Universiti Tenaga Nasional (UNITEN), 43009 Kajang,

Selangor, MALAYSIA {sharif, zaliman, azhana}@uniten.edu.my

ABSTRACT In this paper, an integrated framework for knowledge audit and capture using the task analysis approach is presented. We first identify the types of knowledge that could be ascribed to tasks and analyze a task by breaking it down into subtasks and task elements. We then identify the skills and knowledge required to perform the tasks at this level. The framework is validated and tested on an expertise-based task, the findings of which are compiled as a knowledge-based document.

Keywords Knowledge Management, Knowledge Audit, Knowledge Capture, Task Analysis.

1.0 INTRODUCTION

The problem of knowledge loss is an experience of all organizations, big and small, resulting from the loss of knowledge workers when they leave the organizations. However, most organizations do not find this problem compellingly urgent if failures of performance do not incur high costs to the business as new knowledge workers could be hired to replace the loss or they could provide other measures to avoid serious disruption of business activities. The problem becomes a matter of concern when the loss involves an expert, which may pose serious threats to the business. However, if recurring problems are not addressed, the costs of managing this problem could be astounding. As Soltan (1995) claims, ‘knowledge mismanagement is costing a company far more than managing information systems.’ Earlier attempts of knowledge management (KM) have been corroborated by the deployment of intelligent systems in organizations. Today, such systems, which utilize the AI techniques are coupled with or embedded within information systems and equipped with advanced capabilities (Davenport et al., 2005; Fayyad et al., 1996). The current landscape of knowledge management is epitomized by business and predictive analytics.

Companies that strive to improve their business performance using these data-intensive approaches are competing on optimization-based strategies (Kohavi et al., 2002). In this paper, we present our contribution to KM in an aspect relating to the audit, capture, and reuse of expert knowledge. We propose an integrated framework that concurrently audit and capture knowledge using the task analysis approach. In the next section, we review some of the related work in this research. Section 3.0 discusses the types of knowledge proposed by many researchers and describes our preferred types. In Section 4.0, we present our framework followed by Section 5.0, which describes the framework’s validation process. Section 6.0 highlights the use of the framework’s outcome and Section 7.0 concludes the paper. 2.0 RELATED WORK

The KM process embraces several tasks, which include knowledge creation, collection, organization, dissemination, and maintenance (Awad & Ghaziri, 2004). In this paper, we focus on a small but significant aspect of knowledge collection. The following subsections discuss some of the work that are relevant to our research.

2.1 Task Analysis The task analysis approach to knowledge capture and audit is adapted from the study and analysis of tasks. The purpose of task analysis is to determine the nature of the task, the way in which it is performed, and the behaviors the knowledge worker must exhibit to accomplish the task (Youngman et al., 1986). With task analysis, the types and degrees of knowledge, skills, and abilities the knowledge worker must posses can be determined (Freedman et al., 1982). Usually, the information are used to design training curricula to fulfill the training needs of knowledge workers. While the analysis of tasks carried out for training purposes requires the determination of skills, knowledge and attitudes to

405  

a much greater detail, the task analysis conducted for knowledge capture and audit requires the auditor to determine only the skills and knowledge deficiencies.

2.2 Knowledge Audit There is a consensus among researchers that knowledge audit is the process of identifying the core information and knowledge needs and uses in an organization. It identifies gaps, duplications, flows, and how they contribute to business goals (Thirumoorthy, 2003), (Yelden & Albers, 2004). It also investigates and analyses the current knowledge environment and culminates in a diagnostic and prognostic report on the current corporate ‘knowledge health.’ The audit is, thus, the first major stage in effective knowledge management and corporate knowledge valuation (Hylton, 2002). The importance of knowledge audit is attested by the numerous techniques and methodologies for knowledge audit. Choy et al. (2004), for example, suggest a systematic approach to integrate various knowledge audit related techniques into pre-audit preparation, in-audit process and post-audit analysis. Lauer and Tanniru (2001) propose a methodology to understand the “gaps” in the needs of a knowledge worker. The methodology uses the “process change” research to help build a socio-technical environment critical for knowledge work. Thirumoorthy (2003) proposes a three-step procedure of knowledge audit, which identifies what knowledge currently exists in the targeted area, identifies what knowledge is missing in the targeted area, and provides recommendations to management regarding the status quo and possible improvements to the knowledge management activities in the targeted area.

2.3 Knowledge Capture Knowledge capture or elicitation is a process by which an expert’s thoughts and experiences are captured and documented (Awad & Ghaziri, 2004). Many knowledge capture techniques have been proposed by researchers. However, no one technique can claim superiority over the others. Each of these techniques such as on-site observation; brainstorming; protocol analysis; repertory grid; concept mapping; and nominal group technique is used to capture a particular type of knowledge (Awad & Ghaziri, 2004). Consequently, a knowledge engineer must be able to assess and select a suitable technique or a combination of techniques that ensure the total capture of knowledge from experts. Kingston, Shadbolt and Tate (1996) establish a comprehensive knowledge engineering approach to knowledge-based systems design. The

CommonKADS (Kingston et al., 1996) employs expertise and design models to support knowledge engineers in choosing knowledge representations and programming techniques. The models consist of a three-stage transformation process: application design, architectural design, and platform design. The approach enables useful documentation of system design process and encourage greater modularity and reusability of designs. 3.0 TYPES OF KNOWLEDGE

The knowledge management literature describe two general types of knowledge: tacit and explicit (Nonaka, 1994). Ryle (1969) suggests that knowledge can be classified as “knowing how” and “knowing that.” Others label such knowledge as procedural and declarative knowledge respectively (Awad & Ghaziri, 2004), (Anderson, 1983). Velencei (2003) further details the tacit and explicit types into skills, intuitions and facts. Jorna (2001) prefers a semiotic perspective to knowledge types and considers three types of knowledge: tacit or perceptual knowledge, coded knowledge and theoretical knowledge. Clearly, the various aspects of knowledge make it almost impossible to define specific types of knowledge. We base our framework on the tasks of knowledge workers, not just any tasks but specifically, expertise-based tasks. It does not consider the explicit knowledge assets of an organization, but focus primarily on an expert’s tacit knowledge.

3.1 Development of Framework To develop the framework, we use the task analysis technique while analyzing, identifying and classifying the skills and knowledge into several types. However, the classification is contextual in nature and is derived from the analysis of common organizational tasks. The knowledge types are classified as follows: • motor skill (S), • heuristic (H), • procedural (P), • fundamental (F).

Skill is defined as the ability, talent and craftiness of a knowledge worker to perform a manual task completely and thoroughly. Heuristic refers to the tricks of the trade, rules of thumb, hunches, intuitions, instincts or short cuts which evolve through constant exposure and prolonged experience in a specific task. In this context, cognitive skills are considered as a form of heuristic. We define procedural knowledge as the steps or procedure to perform a task. Knowledge acquired by knowledge workers through formal education and training is the fundamental

knowleand comeducatiothe enfoundatutilize mperformclassify• the

manent

• the of and

• it con

• it wof org

4.0 TH

The intknowleidentifyThe oustructurknowleshows a

Fig

The icomposcaptureidentificwhetherinstrumorganizcharacteauditingconductsubject involve

(a) Ideacc

(b) Deis,

(c) Foknsoutoo

dge. This knomplexity accoronal backgrou

nabling knowtion for knowmore advance

mance of theying the knowlese types of nage and shoities, knowledge tyeach knowled

d hence must bwould make

nvenient, would facilitate

knowledge ganizational go

HE INTEGR

tegrated framedge of a task.

y the tools, proutcome of thire of the taskdge the task a representation

gure 1: The Con

ntegrated frasite processes. The audit prcation of the tr the task can

ments used and zation and docuerize the captug and captuted using the b

matter expes the followingentify the tacomplishment ecompose eachits subtasks,

or each subtnowledge requurces of refereols to be used.

owledge type rding to the qu

und of knowledwledge which wledge seekers

d knowledge eir jobs. Thedge types in thknowledge a

uld be identif

ypes embeddeddge worker aree treated as suc

knowledge

e the alignmentfor the ac

als.

RATED FRAM

ework identifieIt analyzes th

ocedures and ris process is

k, which showis made of. Fn of our concep

ncept of K-Audit

amework cons of knowledrocess is charatype of skills n be performethe references umentation of ure process. Turing the k

brainstorming tert (SME). g activities: asks that conof a job,

h task into its c

task, audit tirements and ences, and the

varies in depualifications andge seekers. It

serves as to acquire an

required for thhe reasons fhis manner areare difficult fied as separa

d within the heae unique to hich, auditing mo

t and integratiochievement

MEWORK

es the skills ane task further references usethe knowledg

ws the skills anFigure 1 belopt.

and Capture

nsists of twdge audit anacterized by thand knowledg

ed, the tools referred to. Th

f the informatioThe processes knowledge atechnique withThe techniqu

ntribute to th

components, th

the skills antheir types, th

e instruments

406

pth nd is a

nd he for e: to

ate

ad im

ore

on of

nd to

ed. ge nd ow

wo nd he ge, or he on of

are h a ue

he

hat

nd he or

If ittaskthe alsoelabstepthe

5.0

To projloca(Ahproj

5.1A bSM

1.

2.

3.

t is possible tok elements, thesources of ref

o be identified boration of the ps to accompliintegrated fram

VALIDATI

validate our fject on the taation for Tehmad et al., 20ject report in th

Figure 2:

Knowledge Abrainstorming

ME, which was The SME versubtasks) perfFor each subtthe following

• describe th• identify th

the knowheuristics fundamen

• whether th• tools and/• references

These querieauditing procinstances, theperformed foinsights and perform the taSteps 1 and 2cable fault loc

o break down e skills and knoferences, and tat this level. Asubtask, which

ish the subtaskmework.

ION OF FRA

framework, weasks associateenaga Nasion010). We preshe following su

: The Integrated

Audit and Capsession was

conducted as frbalized the seformed in eachtask, a facilitat:

he skills and knhe types of k

wledge applied(H), proce

ntal (F) know-lehe subtask couor instrumentss used and refe

es characterizcess for each e SME narrateor a subtask an

finer points task qualitativel2 were repeatedcation.

the subtask inowledge, their the tools used A task elementh shows the dek. Figure 2 dis

AMEWORK

e apply it as aed with cable nal Berhad (sent excerpts oubsections

Framework

pture organized wi

follows: equence of steph task. tor queried the

nowledge applknowledge, whd were skills

edural (P), aedge, ld be performe used, rred to.

ze the knowsubtask. In se

ed detailed actind provided fthat were crucly. d for all the ta

nto its types, could

t is an etailed splays

K

a pilot fault

(TNB) of the

ith an

ps (or

e SME

lied, hether s (S), and/or

ed,

wledge everal ivities

further cial to

sks of

We recforms: AUDII documedocume2010).

5.2 CrThe kncompiliand theskills anwere id As an ithe sam2.1 andThe conTYPESSUB-T The resunion subtaskknowle

JOTR

orded the deta(a) AUDI ((Skills and

ents represeents of Cable F

reation of Knonowledge strucing the skills ae knowledge hnd knowledge

dentified while

llustration, wemple of the knd 2.2 extractedntents of KNOS represent thASKS/TASK

sulting knowledof the know

ks. Abstractingdge structure

OB: Cable FaultTASK REF. NOREF. NO. SUB-T

2.1 For RMswitch (a) Us

sitt

2.2 Connecthe faul

ils of the findi(Task Descripd Knowledgeent the kFault Location

owledge Struccture of a taska task performehe/she needs t

e types (i.e. S, analyzing a su

e show in Tablnowledge strucd from Ahma

OWLEDGE ELe knowledge ELEMENTS

dge structure owledge structug this conceof cable fault

T

t Location .: 2.0

TASKS/TASK E

MU/OLU only, cgear for proper sse insulation testting. High readings

test plugs are ct the test cable olty cable termina

F

ngs in two audptions) and (be). These twknowledge-basen (Ahmad et a

cture k is created ber needs to havto know. TheH, P, and/or F

ubtask.

e 1, a portion cture of Subtasad et al. (2010LEMENTS anstructure of th (Ref. No. 2.1)

of the task is thures of all iept further, tht location is thTable 1: A Samp

TASK: Analy

ELEMENTS

heck test plugs ositting: ter to ensure prop

s indicate that thnot properly sea

of the instrumenal.

Figure 3: Abstra

407

dit b)

wo ed

al.,

by ve se F)

of sk 0). nd he ).

he its he he

unioFigu TherefePro

6.0

ThetypespeMenEdu A kknoknorequmatand

ple of the Knowl

ysis of cable fau

KNOW

of

per

he ated.

• Use I• Judge• Basic

nts to • Basic

ctions of Knowle

on of all the kure 3 illustrate

e knowledge serence documegram (KEP).

KNOWLEDPROGRAM

e KEP segmene (S, H, P, F) icific trainingntoring, On-ucation. Figure

knowledge worowledge gap beowledge structuired by his/hetter of assessin

d practical test

edge Structure

lt

WLEDGE ELEM

Insulation Testere proper sitting (c Electrical Eng.

c circuit theory (F

edge Structures

knowledge strues these abstrac

structure can ent for a Know

DGE ENHANM

nts collections identified fromg programs: -the-Job Traie 4 explains thi

rker undergoesetween his/herture, i.e. the ker job. Identifyng his/her achsessions.

MENTS TYP

r (P) (H) (F)

P, H

F) F

uctures of its ctions.

then be usedwledge Enhance

NCEMENT

of each knowm the framewor

Skills Training or Fs concept.

s a KEP if therr knowledge anknowledge tha

ying this gap is hievement in w

PES Can Perform?

H, F Y

F Y

tasks.

d as a ement

wledge rk into aining, Formal

re is a nd the at are just a

written

 

The skprograminformabecomeuses to

7.0 COW

The intthe steknowlecan be pthe refemeans twould o At thisknowlepracticafield trithe docu In our for docuor discorecurrin

REFE Ahmad

N.a Do

AndersoCo

Figure 4: The K

kills and knom enhance hation in the knoe the new fouperform his/he

ONCLUSIONORK

tegrated frameeps of performdge used and performed, theerences referreto preserve anotherwise be lo

s initial stagedge in naturality and appliials, where botuments in solv

future work, wumenting new overed in the fng problems.

RENCES

d M. S., Rajam., & Tamsir MKnowledge-ba

oc). ProGEM Pon J. R. (1ognition. Harv

K-Enhancement

owledge gainehis/her understowledge-basedund knowledger job.

NS AND FUR

ework presenteming a task, their types, w

e tools or instrued to. The framnd reuse the knost when exper

e, we represenral language icability of thth experts and ving critical pro

we will develoknowledge, w

field while exp

anickam A-S.,M. Z., (2010). A

ased DocumeProject Report,1983). The Avard University

Program

ed out of thtanding of th

d documents ange which he/sh

RTHER

ed here capturthe skills an

whether the tasuments used anmework offers nowledge, whicrts depart.

nt the captureto assess th

he framework non-experts u

oblems.

op a frameworwhich are create

perts solve non

, Omar F., OmA Framework foent System (K, unpublished.Architecture y Press, Boston

408

his he nd he

res nd sk nd a ch

ed he in

use

rk ed n-

mar for Kb-

of n.

Aw

Cho

Dav

Fay

Fre

Hyl

Kin

Koh

Lau

Non

Ryl

Solt

Thi

Vel

Yel

You

wad E. M., GhManagemen

oy S. Y., Lee WSystematic Analysis: IntMapping anProceedings

venport T. H., Competing Education, Report.

yyad U., Piat(1996). FroDiscovery in

edman R. D.(1982). MTheory, ReseSons.

lton A., (20Knowledge-Knowledge A

ngston J., ShaCommon KAPlanning. Pr

havi R., RothleEmerging TCommunicat

uer T. W., TManagemenCase StudInformation

naka I. (19OrganizationOrganization

le G. (1969). Hutchinson &

tan H. (19Knowledge Cranfield Un

irumoorthy P. Chartered A

lencei J. (200Doctus KnProceedingsInformation IASTED, Sc

lden E. F., & ACase For KnKnowledge M

ungman M. BHetwood J. Press.

haziri H. M. t, Prentice HalW. B., & CheuApproach fortegration of Knnd Knowledg

s of I-KNOW ’0Cohen D., & Jon Analytics. Working Kn

etsky-Shapiro om Data Minn Databases. AI., Cooper C.

Management Edearch and Pra

02). MeasuriValue and the Audit. Hylton Aadbolt N. R.,ADS Models foroceedings of Aeder N. J., & STrends in Btions of the ACTanniru M., (t Audit - A

dy. AustralasSystems.

994). A Dynnal Known Science.

The Concept & Co. 995). Invest

Asset, The niversity.

(2003). Knoccountant. 03). An Inte

nowledge-baseds of Internatio

and Knocottsdale. Albers J. A. (2nowledge ManManagement PB., Oxtoby R

(1986). Anal

(2004). Knowll. ung C. F. (200r Knowledge nowledge Invege Flow Ana04. Jacobson A. (2

Babson Execnowledge Res

G., & Smytning to KnowI Magazine. L., Stumpf

ducation: Issuctice, John Wi

ing and AssePivotal Role

Associates. , Tate A., (1or Knowledge-AAAI '96. Simoudis E. (2

Business AnaCM, 45 (8), 45-(2001). Know

Methodologysian Journa

namic Theorledge Cre

t of Mind. Lo

Your CorpReturn is

owledge Audit

lligent Portal d Expert Syonal Conferenowledge Sh

2004). The Buagement. Jour

Practice. R., Monk J. Dlyzing Jobs. G

wledge

04). A Audit

entory, alysis.

2005). cutive search

th P., wledge

S. A. ues in iley &

essing of the

1996). -based

2002). lytics. 48.

wledge y and l of

ry of eation:

ondon:

porate High.

t. The

with ystem. ce on aring,

usiness rnal of

D., & Gower