K-Assessment: Identifying task-relevant knowledge

Scenario Description: The worker can obtain help from experts in conducting the assessment. The worker can access a list of referring tasks. He can arbitrarily click

for further information about a specific task. Finally, after finishing the assessment, he can enter his task workspace to access the relevant knowledge sources provided by the system.

Operations via Interface: The worker can conduct task-assessment to generate his own task profile. If he selects the “assessment” item, the system will guide him to conduct two-phase task-relevance assessment. The worker should give his perceptions of each category. Besides the worker’s perception about the task, he can choose the “expert” column to help him conduct assessment. The results of assessment are submitted to the system’s task profile modeling server to compute the initial task profile. The task profile is expressed as a feature vector of weighted terms.

Figure 19 shows the interface of the two-phase assessment procedure. The detail of collaborative task-relevance assessment is given in Chapter 5.

8.2.3 K-Delivery: Delivering codified knowledge proactively

Scenario Description: Everyone who finished the assessment can enter his task workspace. The system will recommend the task-relevant and the latest information based on the task profile Workers could accept or reject these knowledge items by clicking the feedback web form. Meanwhile, the system will receive feedbacks and modify worker’s profile.

Fig. 18. Interface of K-Processing

Fig. 19. Interface of two-phase assessment

Operations via Interface: The system can proactively deliver task-relevant information based on the worker’s task profiles. Figure 20 shows the top-5 relevant tasks, top-30 relevant documents and 10 task-associated terms provided by the system. A tree-like structure is employed to organize task-relevant information. Once the worker selects a document or a task to read, the detailed information will be displayed, as shown in the right frame of Figure 20. Meanwhile, the worker can view the description of any task-relevant document, as denoted in circle 1. The worker can also conduct feedback on the recommended items. Six relevant degrees are provided by the system- “very low”, “low”, “normal”, “high”, “very high”, and “perfect”, as shown in Figure 21. If the worker gave a positive rating on the knowledge item (document or task sets), the system will preserve the item in the worker’s MyFavorite folder. The detail of disseminating task-relevant knowledge is given in Chapter 6.

8.2.4 K-Sharing: Knowledge support from peer-group

Scenario Description: Once the worker cannot obtain knowledge support from the application of knowledge delivery, he/she can seek the assistance from the application of knowledge sharing. That is, the system identifies peer-groups with similar task needs based on work profiles. The system facilitates knowledge sharing by displaying the shared information such as relevant tasks and documents retrieved

Fig. 20. Interface of knowledge delivery

Fig. 21. Six Degrees of relevance feedback

from peer-group members. Note that all information is calculated timely and automatically according to the feedback results.

Operations via Interface: The system expands the personalized ontology of a worker with the peer-group member’s personalized ontology for knowledge sharing.

Notably, a worker’s personalized ontology represents a worker’s perspective of task-needs on the target task. The personalized ontology is derived from the work profile to record tasks or fields that are relevant to the target task. The left frame of Figure 9 in Section 6.4 shows the sharing tree of “Jia-Yuan Lee”, as denoted in circle 1. A sharing tree is a tree-like structure, which represents the personalized ontology of a worker. Meanwhile, the shared information from task-based peer-groups is also presented in the sharing tree. In the given example, the ontology {H3.3 Information Retrieval and K4.3 Organization Impact, Mining Association Rule for Information Recommendation in Enterprises} is shared from “Mike Lee”, as denoted in circle 2 of Figure 9. Another tree-like structure below the sharing tree is used to organize the

shared document sets from the task-based peer-group (as denoted in circle 3).

Notably, a threshold, α-cut level, which is shown in the top left frame, can be adjusted by the workers to find more peer-group members by decreasing the α value.

The detail of disseminating and sharing task-relevant knowledge is given in Chapter 6.

8.3 Discussions

We also examined the user effort to conduct assessment procedure and the overall useful perception about the proposed system briefly more details are given in the previous publication [86].

User effort: The user effort result showed that the novice workers took 27 minutes on average to complete the procedure, whereas the experienced workers took 16 minutes on average. This result is in line with the research of Marshall and Byrd (1998) that states the use and perception of information system (IS) varies by user groups due to the task domain knowledge. In our post questionnaires, we found that experienced workers seemed satisfied with the design of assessment procedure, whereas the novice users seemed to take more time to conduct the assessment.

Therefore, the collaborative mechanism is especially required for novices.

Satisfaction: For measuring the users’ satisfaction of using the system, a 6-point Likert scale from 1 to 6 was deployed. Two questions were asked after completing the task assessment: one was the difficulty to conduct the task assessment and the other was the usefulness of the recommend items. Table 21 shows the result of the users’ evaluation. Notably, the common consensus about usefulness among experienced workers was low measured by standard deviation (e.g., the value is higher than that of novices). This resulted from the fact that some of experienced workers were not satisfied with the quantity of knowledge support. Increasing system scalability is a major task in our future work.

Finally, according to our post questionnaires about the proposed system, the novices reflect that they want to get more help from humane resource, whereas the experienced workers need more number of task-relevant knowledge supports. It indicates that the sharing mechanism is more essential for novices than that of experienced workers. On the other hand, the web mining technique is more desirable for experienced workers for acquiring more task-relevant knowledge form the Web pages.

Table

21.

Useful to support task 4.200 0.448 4.400 1.342