To discover early an abnormality of an apparatus, concerning an apparatus deterioration evaluation support method and device.

The apparatus deterioration evaluation support device 1 measures and stores initial acoustic data (S401). The device 1 measures and stores periodically present acoustic data (S402). The device 1 compares the initial acoustic data with the present acoustic data, extracts a different pattern, and stores it as a candidate of a symptom pattern (S403). If the candidate is not extracted (NO in S404), the device 1 remeasures the present acoustic data (S402). If the candidate is extracted (YES in S404), the device 1 sets a discrimination function for discriminating a normal pattern and a candidate, by using a support vector machine (S405). The device 1 grasps a changing tendency in the present acoustic data having a generation frequency of the same pattern as the candidate, by using the discrimination function (s406). If the generation frequency has an increasing tendency (YES in S407), The device 1 stores the candidate as the symptom pattern (S408). The device 1 grasps in detail the generation frequency of the symptom pattern (S409), and evaluates a deterioration state of an observation object apparatus (S410).