Subject-Oriented Semantic Knowledge Warehouse (SSKW) to Support Cognitive DSS

摘要

The communication between cognitive DSS and data warehouse tends to be inefficient due to their contradictory knowledge/data oriented nature. Data-to-knowledge conversion requires specialized techniques, whereas knowledge-to-data conversion results in loss of knowledge. To address these issues, a subject-oriented semantic knowledge warehouse (SSKW) is proposed, to provide relevant and precise knowledge to CDSS. The SSKW consists of: a) object/process/event/relationship (OPER) model to store domain knowledge in a unified fashion; and, b) a subjective view database, containing opinions of stakeholders about various OPER knowledge elements. A case study to compare the performance of the SSKW-based CDSS against a DW-based CDSS is presented. The results show that SSKW improves communication efficiency, provides relevant and precise domain knowledge to CDSS in less decision cycles, minimizes the loss of knowledge, and helps decision maker to quickly grasp the decision situation through its human-centric nature.