Server apparatus, odor sensor data analysis method, and computer readable recording medium for unfixed odor analysis targets

An odor sensor data analysis method that uses a terminal apparatus for collecting sensor data from a Membrane-type Surface Stress (MSS) odor sensor that outputs the sensor data in reaction to a plurality of types of odors and a server apparatus configured to be communicably connected to the terminal apparatus, the method comprising:

• with the terminal apparatus, collecting the sensor data;
• with the server apparatus, selecting an analysis model capable of analyzing a designated odor analysis target, from among a plurality of analysis models created by machine learning for analyzing specific odor analysis targets by analyzing the sensor data, and determining preprocessing to be performed on the sensor data, according to the selected analysis model;
• with the terminal apparatus, executing the determined preprocessing on the sensor data;
• with the terminal apparatus, transmitting the preprocessed sensor data to the server apparatus;
• with the server apparatus, executing analysis processing of the designated odor analysis target, by applying the selected analysis model to the preprocessed sensor data transmitted thereto from the terminal apparatus;
• with the server apparatus, transmitting information indicating a result of the analysis processing to the terminal apparatus;
• and
• with the terminal apparatus, holding the information indicating the result of the analysis processing.

A non-transitory computer-readable recording medium that includes a program recorded thereon, the program including instructions that cause a computer configured to be communicably connected to a terminal apparatus for collecting sensor data from a Membrane-type Surface Stress (MSS) odor sensor that outputs the sensor data in reaction to a plurality of types of odors to carry out:

• holding a plurality of analysis models created by machine learning for analyzing specific odor analysis targets, based on the sensor data;
• selecting an analysis model capable of analyzing a designated odor analysis target, from among the plurality of analysis models, determining preprocessing to be performed on the sensor data, according to the selected analysis model, and causing the terminal apparatus to execute the determined preprocessing on the sensor data;
• upon the preprocessed sensor data being transmitted thereto from the terminal apparatus, executing analysis processing of the designated odor analysis target, by applying the selected analysis model to the preprocessed sensor data transmitted thereto;
• and
• transmitting information indicating a result of the analysis processing to the terminal apparatus.

A non-transitory computer-readable recording medium that includes a program recorded thereon, the program including instructions that cause a computer for collecting sensor data from a Membrane-type Surface Stress (MSS) odor sensor that outputs the sensor data in reaction to a plurality of types of odors and configured to be communicably connected to a server apparatus that holds a plurality of analysis models created by machine learning for analyzing specific odor analysis targets based on the sensor data to carry out:

• collecting the sensor data;
• in a case where, in the server apparatus, an analysis model capable of analyzing a designated odor analysis target is selected, from among the plurality of analysis models, and preprocessing to be performed on the sensor data is determined, according to the selected analysis model, executing the determined preprocessing on the collected sensor data;
• transmitting the preprocessed sensor data to the server apparatus;
• and
• in a case where, in the server apparatus, analysis processing of the designated analysis target is executed, by the selected analysis model being applied to the preprocessed sensor data transmitted thereto, holding information indicating a result of the analysis processing.