Observation information processing apparatus, non-transitory computer-readable recording medium, and observation information processing method

A non-transitory computer-readable recording medium having stored therein a program for causing a computer to execute a observation information process comprising:

• calculating, for each mesh, a support that is a number of records for which a condition part of an association rule becomes true in records included in a mesh, based on meshes each associated with records each including observation information having information identifying a geographical position and an event that occurs at the geographical position, the information being associated with the event;
• calculating, for each mesh, a confidence that is a ratio of records for which a conclusion part of the association rule also becomes true, to the records for which the condition part becomes true;
• generating an adjacent mesh set by merging adjacent ones of the meshes;
• calculating a confidence for each adjacent mesh based on the support and confidence of each mesh included in the adjacent mesh set;
• setting the smallest one of the confidences calculated as a new confidence threshold value;
• calculating, for each mesh in the adjacent mesh, a first value by multiplying the support and the confidence;
• calculating, for each mesh, in the adjacent mesh, a second value by dividing the first value by a minimum support among the supports of the meshes;
• excluding first meshes included in the adjacent mesh set from the adjacent mesh set, the first meshes having the second value which is smaller than the confidence threshold value;
• and repeatedly executing each process.

An observation information processing method executed by a computer, the observation information processing method comprising:

• calculating, by a processor, for each mesh, a support that is a number of records for which a condition part of an association rule becomes true in records included in a mesh, based on meshes each associated with records each including observation information having information identifying a geographical position and an event that occurs at the geographical position, the information being associated with the event;
• calculating, by the processor, for each mesh, a confidence that is a ratio of records for which a conclusion part of the association rule also becomes true, to the records for which the condition part becomes true;
• generating, by the processor, an adjacent mesh set by merging adjacent ones of the meshes;
• calculating, by the processor, a confidence for each adjacent mesh based on the support and confidence of each mesh included in the adjacent mesh set;
• setting, by the processor, the smallest one of the confidences calculated as a new confidence threshold value;
• calculating, by the processor, for each mesh in the adjacent mesh, a first value by multiplying the support and the confidence;
• calculating, by the processor, for each mesh, in the adjacent mesh, a second value by dividing the first value by a minimum support among the supports of the meshes;
• excluding, by the processor, first meshes included in the adjacent mesh set from the adjacent mesh set, the first meshes having the second value which is smaller than the confidence threshold value;
• and repeatedly executing each process.