For a given operation, different roles may require different degrees of knowledge in different knowledge categories. A role-operation knowledge requirement degree profile is a set of 4-tuple <role r, operation op, knowledge category ca, knowledge requirement degree krdeg>. Each tuple indicates that the degree of knowledge required by role r with respect to operation op in knowledge category ca is krdeg. The knowledge requirement degree is used to define how specific or general the required knowledge in a certain knowledge category should be for role r while executing operation op.
Given a sourcing department manager role r and two operations related to the part sourcing task in Figure 2: evaluating display suppliers (op1) and surveying battery suppliers (op2), Figure 18 shows a role-based framework applied in the context. The role-based framework describes the relationships among role, operation and knowledge node which are the basic concepts for generating a role-based virtual KF. Overall speaking, two phases are required to generate virtual KFs. The two phases are mutually independent.
Phase I generates virtual KNs and Phase II derives knowledge concepts for these virtual KNs. In Figure 18, the role-operation relevance profile, base knowledge node profile and threshold TH are used to generate virtual KNs in Phase I; the role-operation knowledge requirement degree profile and operation required knowledge concept profile are utilized to abstract knowledge concepts for virtual KNs in Phase II.
In this example, the relevance degrees ordeg of r to operation op1 and op2 are 0.2 and 0.05 respectively (shown in role-operation relevance profile); op1 and op2 is associated with kn1 (shown in base knowledge node profile); the knowledge concepts required in the marketing category to perform op1 are {c131, c132} and in the hardware category is {c621}
(shown in operation required knowledge concept profile); the knowledge requirement degree krdeg of role r to perform op1 is 0.8 for the marketing category and 1.0 for the
48
Knowledge requirement degree to abstract knowledge concept Role- KN relevance degree to generate virtual KN
Threshold TH = 0.4
hardware category (shown in role-operation knowledge requirement degree profile). The above information is used to generate a role-based virtual KF as well as derive the corresponding knowledge concepts for the virtual KNs in the virtual KF.
Figure 18. A role-based framework with examples.
5.2.1 Construction of role-operation relevance profile
Initially, role-operation relevance profiles are constructed by analyzing a role’s operating logs. Let T denote the number of times that role r performs all of the assigned operations in a period, and let N denote the number of times that role r performs an operation op in the same period. The default operation relevance degree of r with respect to op is N/T. Take Figure 18 as an example, role r performs op1 four times, op2 once, and other operations 15 times; then T equals 20. The operation relevance degree ordeg of r is 0.2 for op1 and 0.05 for op2 as shown in Figure 18. Intuitively, a higher ordeg indicates greater relevance between a role and an operation. Moreover, the higher the cost associated with an operation performed by a role, the higher the relevance degree of the operation to the role will be. Let Q denote the total cost of the operations assigned to role r, and let C denote the cost of a specific operation op. The ordeg of role r when performing operation op is C/Q. Based on activity-based costing (ABC) models, the cost can be measured in terms of the time and resources expended by roles when they perform assigned operations.
In summary, different statistics can be extracted from the historical log data. Some
49
decision-making methods can be employed to derive the ordeg by combining the statistics.
It is noteworthy, while implementing a role-based KFV system based on the r-KFV model, that the proposed methods and parameters for constructing the role-operation relevance profiles need to be fine-tuned in terms of the culture of organizations, the accommodation of peripheral systems and the context of operations. Workers’ future operation logs can also be used to adjust ordeg to satisfy their real time knowledge-needs.
These adjustments are essential to obtain appropriate virtual knowledge flows.
5.2.2 Evaluation of role-knowledge node relevance
The relevance degree between a given role and a base knowledge node can be derived from the role-operation relevance profiles and base knowledge node profiles. For example, role-operation relevance profile {< r, op1, 0.2 >, < r, op2, 0.05 > …} and a knowledge node profile {< kn1, op1>, < kn1, op2>, …}, the relevance degree between role r and base knowledge node kn1 is max (0.2, 0.05, …) = 0.2. That is, the relevance degree between role r and base knowledge node kn1 is the maximum of the ordeg of multiple operations that r is authorized to refer to base knowledge node kn1.
5.3 Procedures for deriving role-based virtual knowledge flows
This section presents the approach for discovering role-based virtual KFs suitable for participating roles through applying the role-based framework. The approach involves three steps: (a) identifying role-based virtual KNs by aggregating base KNs based on their relevance to a role, (b) building virtual dependencies to connect the identified virtual KNs, and (c) deriving knowledge concepts of the identified virtual KNs based on the operation required knowledge concepts and role’s operation knowledge requirement degrees.
5.3.1 Identifying role-based virtual knowledge nodes
The identification procedure in Figure 19 starts from taking the highest ordering base KNs from a base KF as a seed node. Beginning with the seed, the procedure repeatedly aggregate the adjacent base KNs of the seed according to the descending order of their
50
role-knowledge node relevance degrees. Until the total relevance degree of the aggregated base KNs approximates a granular threshold, a virtual KN is identified. Other base KNs which are not aggregated forms a residual knowledge node set (RKNS). The base KNs in RKNS mean that they are not the member knowledge nodes of any virtual KNs yet. The procedure takes the highest ordering base KN from RKNS as another seed node and repeats another aggregation run. The loop continues until no more base KNs in RKNS. Hence, all role-based virtual KNs have been generated. Then, the virtual dependencies among the identified virtual KNs can be set as described in Section 4.4.2.
Moreover, during the aggregating loop, the order-preserving property should be checked. The order-preserving approach discussed in Section 4.1.1 is applied to ensure that a role-based virtual KF maintains the knowledge referencing order in it as the order in the corresponding base KF. The detail algorithm of identifying role-based virtual KNs is described in a previous work [35].
procedure Identify Virtual KN (input: base KF, role r, threshold TH, output: virtual KF) begin
put waiting-for-aggregated base KNs to RKNS (residual KN set) repeat
select the highest order base KN x from RKNS
/* generate a role-based virtual KN which consists of x */
add a neighbor of x with the largest relevance degree to form temp virtual KN while (temp virtual KN statisfy
Order-preserving
FTRD (r, temp virtual KN) <= TH No overlap base KN)
do (greedily include additional neighboring KN with the largest relevance degree and apply order-preserving approach to add relevant KNs to form temp virtual KN)
generate a role-based virtual KN according to prior temp virtual KN put the generated virtual KN in virtual KF
until no more base KNs in RKNS
51
set virtual dependencies for all identified virtual KNs end
Figure 19. Procedure for identifying role-based virtual KNs.