ABSTRACT

A Sense and Respond (SaR) system endows a Business Intelligence system with the intelligence needed to react timely to exogenous as well as endogenous events. To this end, a SaR system needs to know the Key Performance Indicators (KPIs) that must be maximized as well as their relative weights. While the first information can be easily obtained through interviews, the second one is quite hard to get. This motivates the investigation of methods and tools to address this problem.

In such a context, the main contributions of this paper are the following. First, we show how KPIs can be effectively defined using linear constraints. Second, we show how the problem of computing the actions that the SaR system proposes to the manager can be formulated as a Mixed Integer Linear Programming (MILP) problem. Third, we show how KPI weights can be computed from previous managing decisions by solving a suitable MILP problem. Fourth, we provide experimental results showing the effectiveness of the proposed approach.

摘要

感知和响应（SaR）系统为商务智能系统提供及时响应于外部事件和内部事件所需的智能。为此，SaR系统需要了解必须最大化的关键绩效指标（KPI）及其相对权重。虽然可以通过访谈轻松获得第一个信息，但是很难获得第二个信息。这激发了研究解决该问题的方法和工具的动力。

在这种情况下，本文的主要贡献如下。首先，我们展示如何使用线性约束有效地定义KPI 。其次，我们展示了如何计算SaR系统向管理者提出的动作的问题可以表述为混合整数线性规划（MILP）问题。第三，我们展示了如何通过解决合适的MILP问题从先前的管理决策中计算出KPI权重。第四，我们提供的实验结果表明了该方法的有效性。