Seismic observation device, seismic observation method, and recording medium on which seismic observation program is recorded

A seismic observation method performed by at least one processor of a seismic observation device, the method comprising:

• receiving input of time-series data of measurement values of a vibration from one or more sensors;
• retrieving a trained model that is trained according to machine learning;
• analyzing, by the trained model, the time-series data to determine a time period of the time-series data that is a processing;
• and
• detecting, during the time period, a trigger indicating a vibration different from a normal vibration;
• and
• acquiring, based upon detection of the trigger, a likelihood of classifying a cause of the vibration indicated in the time-series data in the time period, wherein the cause is classified as at least one of a near-field earthquake, a far-field earthquake, a low frequency earthquake, an artificial earthquake, a micro-earthquake, and a volcanic earthquake.

A non-transitory computer readable-medium having instructions stored therein, which when executed by a processor in a seismic observation device cause the processor to:

• receive input of time-series data of measurement values of a vibration from one or more sensors;
• retrieve a trained model that is trained according to machine learning;
• analyze, by the trained model, the time-series data to determine a time period of the time-series data that is a processing target;
• detect, during the time period, a trigger indicating a vibration different from a normal vibration;
• and
• acquire, based upon detection of the trigger, a likelihood of classifying a cause of the vibration indicated in the time-series data in the time period, wherein the cause is classified as at least one of a near-field earthquake, a far-field earthquake, a low frequency earthquake, an artificial earthquake, a micro-earthquake, and a volcanic earthquake.