Mining knowledge flow for modeling the information needs of task-based groups

Abstract

Knowledge is the most important resource to create core competitive advantages for an organization. Such knowledge is circulated and accumulated by a knowledge flow (KF) in an organization to support worker tasks. Workers may cooperate and participate in several task-based groups to fulfill their needs. In this paper, we propose a group-based knowledge flow mining algorithm which integrates information retrieval and data mining techniques for mining and constructing the group-based KF (GKF) for task-based groups. The GKF is expressed as a directed knowledge graph to represent the knowledge referencing behavior for a group of workers with similar task needs. The frequent knowledge referencing paths are identified from the knowledge graph to indicate the frequent knowledge flows of the workers. We also implement a prototype of GKF mining system to demonstrate the effectiveness of our proposed method.