Using case-based reasoning to retrieve incident reports

Incident reporting schemes are being introduced into many industries. This growth is creating new challenges. In particular, it can be difficult to spot emerging trends and common features amongst the thousands of reports that are submitted to many national and international schemes. Traditional databases offer little support here because query formation often defeats even relatively skilled analysts. Similarly, free-text search engines have technical limitations that make it difficult to identify incidents in which certain causes were NOT a factor. As a result many of these systems yield results that have poor precision and low recall values. This paper, therefore, argues that alternative techniques must be developed to support the indexing and retrieval of similar cases from within large-scale incident reporting schemes. In particular, we show how the application of the US Navy's Conversational Decision Aids Environment (NaCoDAE) can be extended from the domain of decision support to help analysts retrieve information about previous incidents. We have validated this approach by applying it to two data-sets from the US Aviation Safety Reporting System (ASRS).