A case-based reasoning tool to recommend drinking water source protection actions

In response to a need expressed by water actors to support them in identifying appropriate actions for protecting drinking water sources, we modelled, designed, implemented, tested, and validated a prototype case-based reasoning (CBR) system. This required an intensive knowledge acquisition and structuring process which we conducted in collaboration with 102 water management and governance actors in Quebec, Canada. Knowledge was organized in a case base containing nearly 200 past experiences implementing water protection actions at different decision-making scales (local, regional, provincial), by various organizations (government, municipalities, water-related associations), over the last two decades. This paper describes our case engineering process to design case attributes based on content analysis. It presents the case edition, the case retrieval process, and the prototype's implementation and validation, conducted through a rigorous and transparent participatory procedure. This is, to our knowledge, the first successful attempt to apply a CBR approach to support water actors in protecting drinking water sources. It provides empirical evidence of the positive potential of such an approach for knowledge sharing and transfer and can be generalized to other similar contexts.