A knowledge model-driven recommender system for business transformation

Engineering approach for transforming business process and workforce management is an emerging area that is gaining increasing interest for its many promising applications in service science. Yet, three major challenges stand out in engineering any business transformation process. The first challenge is accurate modeling of the transformation knowledge that may be scattered across multiple application domains. The second challenge is efficient and precise real-time evaluation based on the knowledge model. The third challenge will be to generate intelligent recommendations about business transformation that can significantly improve the business process and drive down opportunity costs. This paper proposes an integrated knowledge model-driven recommender system that effectively addresses all of the three abovementioned challenges of modeling, evaluating, and recommending. Using a real-world case study on warranty processing at a major automotive manufacturer, this paper presents a novel business transformation framework that consists of a knowledge model on business value drivers and metrics, an evaluation engine for processing real-time business events, and a recommendation engine that utilizes information obtained from the evaluation engine to suggest new processes and workforce allocation strategy, which can be subjected to a new cycle of modeling and evaluation to complete a feedback loop. Our experimental study using the real-world data results in a "25/75" rule in predictive warranty data processing: 25% of information contains 75% of business information entropy, thereby demonstrating the effectiveness of the system.