Proposal system, proposal method, and proposal program
A machine learning-based prediction model quickly identifies countermeasures for microorganisms by analyzing spectral data, addressing the slow response of existing methods and providing timely recommendations.
Patent Information
- Application Number
- JP2024008014
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2024-01-23
- Publication Date
- 2025-08-04
AI Technical Summary
Existing methods are slow in proposing effective countermeasures against microorganisms, often taking more than half a month to provide recommendations.
A computer system using a prediction model trained through machine learning to analyze spectral data of microorganisms, enabling rapid determination of recommended countermeasures based on spectral data input.
Enables quick proposal of countermeasures against microorganisms, potentially on the same day, by leveraging a trained prediction model to process spectral data and determine appropriate actions.
Smart Images
Figure 2025113721000001_ABST
Abstract
Description
Technical Field
[0001] One aspect of the present disclosure relates to a proposal system, a proposal method, and a proposal program.
Background Art
[0002] Non-Patent Document 1 describes label-free Raman spectroscopy for monitoring the types of resistance and the mechanisms of acquired resistance in a bacterial population of Escherichia coli in the absence of antibiotics.
Prior Art Documents
Non-Patent Documents
[0003]
Non-Patent Document 1
Summary of the Invention
Problems to be Solved by the Invention
[0004] It is desired to quickly propose countermeasures against microorganisms.
Means for Solving the Problems
[0005] The proposed system according to one aspect of the present disclosure includes at least one processor. The at least one processor acquires target spectral data obtained by analyzing a target microorganism, inputs the target spectral data into a prediction model learned to perform a prediction based on the spectral data of the microorganism, obtains a prediction result regarding the target microorganism, and based on the prediction result, determines at least one of a plurality of candidate countermeasures preset as countermeasures for the microorganism as a recommended countermeasure for the target microorganism.
[0006] The proposed method according to one aspect of the present disclosure is executed by a proposed system including at least one processor. This proposed method includes the steps of acquiring target spectral data obtained by analyzing a target microorganism, inputting the target spectral data into a prediction model learned to perform a prediction based on the spectral data of the microorganism, obtaining a prediction result regarding the target microorganism, and based on the prediction result, determining at least one of a plurality of candidate countermeasures preset as countermeasures for the microorganism as a recommended countermeasure for the target microorganism.
[0007] The proposed program according to one aspect of the present disclosure causes a computer to execute the steps of acquiring target spectral data obtained by analyzing a target microorganism, inputting the target spectral data into a prediction model learned to perform a prediction based on the spectral data of the microorganism, obtaining a prediction result regarding the target microorganism, and based on the prediction result, determining at least one of a plurality of candidate countermeasures preset as countermeasures for the microorganism as a recommended countermeasure for the target microorganism.
[0008] In such an aspect, based on the prediction result obtained by inputting the target spectral data of the target microorganism into the prediction model obtained by learning, the countermeasure for the target microorganism is determined from a plurality of candidate countermeasures. By introducing the prediction model, as long as the spectral data of the microorganism is prepared, a countermeasure for the microorganism can be quickly proposed.
Advantages of the Invention
[0009] According to one aspect of the present disclosure, countermeasures against microorganisms can be quickly proposed.
Brief Description of the Drawings
[0010]
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Modes for Carrying Out the Invention
[0011] Hereinafter, various examples in the present disclosure will be described in detail with reference to the accompanying drawings. In the description of the drawings, the same or equivalent elements are denoted by the same reference numerals, and redundant descriptions are omitted.
[0012] [Overview of the System] The proposal system according to the present disclosure is a computer system that proposes countermeasures against microorganisms such as bacteria to a user. Countermeasures against microorganisms refer to methods for avoiding or suppressing the growth of microorganisms. In one example, the proposal system proposes countermeasures against undesirable microorganisms contained in food and drink. In the present disclosure, the microorganisms to which countermeasures are proposed are also referred to as "target microorganisms". Also, the countermeasures proposed to the user are also referred to as "recommended countermeasures".
[0013] The proposed system processes target spectral data, which is spectral data obtained by analyzing a target microorganism, and determines recommended countermeasures based on the result of the processing. In one example, the proposed system uses a prediction model to process the target spectral data. The prediction model is a trained model generated by machine learning, which is a method of autonomously finding laws or rules by repeatedly learning based on given information. The prediction model is constructed using an algorithm and a data structure. By using the prediction model, it is possible to quickly propose countermeasures for a microorganism as long as the spectral data of the microorganism is prepared. For example, it becomes possible to propose countermeasures that conventionally took more than half a month on the same day.
[0014] The proposed system may further use a target category, which is a food and beverage category corresponding to the target microorganism, to determine the recommended countermeasures. The food and beverage category corresponding to a microorganism refers to the food and beverage that is the target for implementing countermeasures against the microorganism.
[0015] The countermeasures proposed in the proposed system may be countermeasures regarding additives such as sodium acetate, glycine, and sorbic acid, or may be countermeasures regarding manufacturing processes such as heat treatment, anaerobic packaging, low pH, and low temperature control.
[0016] The analysis method of the target microorganism for obtaining the spectral data may be appropriately selected. For example, the spectral data processed by the proposed system can be one of various spectra such as Raman spectra, mass spectra, and infrared spectra.
[0017] [Configuration of the System] FIG. 1 is a diagram showing the functional configuration of a proposed system 10 according to an example. In this example, the proposed system 10 is connected to a learning database 21, a countermeasure database 22, and a user terminal 30 via a communication network such as the Internet or an intranet.
[0018] In one example, the proposed system 10 includes, as functional components, a learning unit 11, an acquisition unit 12, a first prediction unit 13, a second prediction unit 14, a search unit 15, a filtering unit 16, and an output unit 17. The learning unit 11 is a functional module that trains a prediction model 40 to perform predictions based on spectral data of microorganisms. In one example, the learning unit 11 generates a countermeasure prediction model 41 and a bacterial species prediction model 42 as the prediction model 40 by machine learning. The learning unit 11 accesses a learning database 21 for the machine learning. The acquisition unit 12 is a functional module that acquires target data regarding a target microorganism. The target data may include target spectral data and a target category. The first prediction unit 13 is a functional module that predicts countermeasures from the target spectral data using the countermeasure prediction model 41. The second prediction unit 14 is a functional module that predicts countermeasures from the target spectral data using the bacterial species prediction model 42 and a countermeasure database 22. The search unit 15 is a functional module that searches the countermeasure database 22 for countermeasures corresponding to the target category. The filtering unit 16 is a functional module that determines recommended countermeasures based on a set of countermeasures obtained by at least one of the first prediction unit 13, the second prediction unit 14, and the search unit 15. The output unit 17 is a functional module that outputs the recommended countermeasures for proposal to the user.
[0019] FIG. 2 is a diagram showing an example of the hardware configuration of a computer used in the proposed system 10. In one example, the computer 100 includes a processor 101, a main memory unit 102, an auxiliary storage unit 103, a communication control unit 104, an input device 105, and an output device 106.
[0020] The processor 101 is an arithmetic unit that executes an operating system and application programs. Examples of the processor include a CPU (Central Processing Unit) and a GPU (Graphics Processing Unit).
[0021] The main memory unit 102 is a device that stores programs for operating the proposal system 10, calculation results output from the processor 101, and the like. The main memory unit 102 is constituted by at least one of, for example, a ROM (Read Only Memory) and a RAM (Random Access Memory).
[0022] The auxiliary storage unit 103 is a device that can generally store a larger amount of data than the main memory unit 102. The auxiliary storage unit 103 is constituted by a non-volatile storage medium such as a hard disk or a flash memory. The auxiliary storage unit 103 stores a proposal program 110 for causing at least one computer to function as the proposal system 10 and various types of data.
[0023] The communication control unit 104 is a device that executes data communication with other computers via a communication network. The communication control unit 104 is constituted by, for example, a network card or a wireless communication module.
[0024] The input device 105 is a device that receives data based on operations or actions of a user who manages the proposal system 10. The input device 105 is constituted by, for example, a keyboard, a mouse, a touch panel, or the like.
[0025] The output device 106 is a device that outputs data processed by the proposal system 10. The output device 106 is constituted by, for example, a monitor and a speaker.
[0026] The proposal system 10 may be constituted by one computer 100, or may be constituted by a plurality of computers 100. When a plurality of computers 100 are used, these computers 100 are connected via a communication network such as the Internet or an intranet, thereby logically constructing one proposal system 10.
[0027] Each functional module of the proposal system 10 is realized by causing the processor 101 or the main memory unit 102 to load the proposal program 110 and execute the proposal program 110. The proposal program 110 includes code for realizing each functional module of the proposal system 10. The processor 101 operates the communication control unit 104, the input device 105, or the output device 106 according to the proposal program 110, and reads and writes data in the main memory unit 102 or the auxiliary storage unit 103. Data or databases required for processing may be stored in the main memory unit 102 or the auxiliary storage unit 103.
[0028] The proposal program 110 may be provided after being non-temporarily recorded on a tangible recording medium such as a CD-ROM, a DVD-ROM, or a semiconductor memory. Alternatively, the proposal program 110 may be provided via a communication network as a data signal superimposed on a carrier wave. The provided proposal program 110 is stored in the auxiliary storage unit 103.
[0029] The learning database 21 is a device that stores teacher data used for machine learning. FIG. 3 is a diagram showing an example of the teacher data stored in the learning database 21. In this example, the learning database 21 stores first teacher data 21a for generating the countermeasure prediction model 41 and second teacher data 21b for generating the bacterial species prediction model 42.
[0030] Each data record of the first teacher data 21a includes a sample microorganism ID, spectrum information, and countermeasure information. The sample microorganism ID is an identifier that uniquely identifies a sample microorganism, which is the microorganism prepared as a sample. The spectrum information is information indicating the spectrum obtained by analyzing the sample microorganism, and is represented by, for example, the intensity at each of n wavelengths. The countermeasure information is information indicating candidate countermeasures for the sample microorganism. For example, each of one or more candidate countermeasures indicates, in binary, whether it is effective for the sample microorganism. In the example of FIG. 3, the value "1" indicates that the candidate countermeasure is effective for the sample microorganism, and the value "0" indicates that the candidate countermeasure is not effective for the sample microorganism. The countermeasure information is used as the correct answer in machine learning.
[0031] Each data record of the second teacher data 21b includes a sample microorganism ID, spectrum information, and the bacterial species of the sample microorganism. The bacterial species is used as the correct answer in machine learning.
[0032] The countermeasure database 22 is a device that stores countermeasure data indicating candidate countermeasures prepared in advance as candidate recommended countermeasures. FIG. 4 is a diagram showing an example of the countermeasure data stored in the countermeasure database 22. In this example, the countermeasure database 22 stores first countermeasure data 22a indicating the correspondence between a plurality of bacterial species and a plurality of candidate countermeasures, and second countermeasure data 22b indicating the correspondence between a plurality of food and drink categories and a plurality of candidate countermeasures.
[0033] Each data record of the first countermeasure data 22a includes a bacterial species and countermeasure information. The countermeasure information is information indicating candidate countermeasures for the bacterial species. For example, each of one or more candidate countermeasures indicates, in binary, whether it is effective for the bacterial species. In the example of FIG. 4, the value "1" indicates that the candidate countermeasure is effective for the bacterial species, and the value "0" indicates that the candidate countermeasure is not effective for the bacterial species.
[0034] Each data record of the second countermeasure data 22b includes a food and drink category and countermeasure information. The countermeasure information is information indicating candidate countermeasures in the food and drink category. For example, each of one or more candidate countermeasures indicates in binary whether it is effective in the food and drink category. In the example of FIG. 4, the value “1” indicates that the candidate countermeasure is effective in the food and drink category, and the value “0” indicates that the candidate countermeasure is not effective in the food and drink category. In the example of FIG. 4, since heat treatment is not suitable for raw vegetables, “0” is set for heat treatment for raw vegetables. Since low-temperature management is not suitable for bread, “0” is set for low-temperature management for bread.
[0035] As described above, the countermeasure information indicated by the first teacher data 21a, the first countermeasure data 22a, and the second countermeasure data 22b may indicate candidate countermeasures regarding additives or may indicate candidate countermeasures regarding manufacturing processes. The candidate countermeasures indicated by the countermeasure information may be unified among the first teacher data 21a, the first countermeasure data 22a, and the second countermeasure data 22b. Alternatively, at least a part of the candidate countermeasures indicated by the countermeasure information may be different among these data.
[0036] The user terminal 30 is a computer used by a user of the proposal system 10. The user terminal 30 can be various computers such as a personal computer, a workstation, a tablet terminal, a smartphone, and a wearable terminal.
[0037] [Operation of the System] The processing executed by the proposal system 10 includes generation of the prediction model 40 and proposal of countermeasures using the prediction model 40. Generation of the prediction model 40 corresponds to the learning phase of machine learning. Proposal of countermeasures corresponds to the operation phase or prediction phase of machine learning.
[0038] [Generation of Prediction Model] The learning unit 11 generates a countermeasure prediction model 41 through machine learning using the first teacher data 21a. The learning unit 11 executes the following processing for each data record of the first teacher data 21a. That is, the learning unit 11 inputs the spectrum information indicated by the data record into the machine learning model. The learning unit 11 may create, select, or extract feature quantities from the spectrum information by means of principal component analysis, genetic algorithms, etc., and then input the feature quantities into the machine learning model. The learning unit 11 updates the parameter group in the machine learning model by means of processing such as backpropagation (error backpropagation method) based on the error between the countermeasure information predicted by the machine learning model and the countermeasure information (correct answer) indicated by the data record. The learning unit 11 repeats the processing for each data record until a predetermined end condition is satisfied to generate the countermeasure prediction model 41. The end condition may be to process all the data records of the first teacher data 21a. Through such machine learning, the learning unit 11 learns the countermeasure prediction model 41 so as to receive the spectrum data of the microorganism and predict one or more countermeasures against the microorganism.
[0039] The learning unit 11 generates a bacterial species prediction model 42 through machine learning using the second teacher data 21b. The learning unit 11 executes the following processing for each data record of the second teacher data 21b. That is, the learning unit 11 inputs the spectrum information indicated by the data record into the machine learning model. The learning unit 11 may create, select, or extract feature quantities from the spectrum information by means of principal component analysis, genetic algorithms, etc., and then input the feature quantities into the machine learning model. The learning unit 11 updates the parameter group in the machine learning model by means of processing such as backpropagation (error backpropagation method) based on the error between the bacterial species predicted by the machine learning model and the bacterial species (correct answer) indicated by the data record. The learning unit 11 repeats the processing for each data record until a predetermined end condition is satisfied to generate the bacterial species prediction model 42. The end condition may be to process all the data records of the second teacher data 21b. Through such machine learning, the learning unit 11 learns the bacterial species prediction model 42 so as to receive the spectrum data of the microorganism and predict the bacterial species of the microorganism.
[0040] Examples of machine learning for generating the prediction model 40 include a support vector machine (SVM), a neural network, a random forest, and deep learning. The specific method of machine learning may be the same or different between the countermeasure prediction model 41 and the bacterial species prediction model 42.
[0041] Note that both the countermeasure prediction model 41 and the bacterial species prediction model 42 are computational models estimated to be optimal, but are not necessarily "computational models that are actually optimal".
[0042] (Proposal of countermeasures) With reference to FIG. 5, an example of the proposal process executed by the proposal system 10 will be described, and an example of the proposal method according to the present disclosure will be described. FIG. 5 is a flowchart showing the example as a processing flow S1.
[0043] In step S11, the acquisition unit 12 acquires target data regarding the target microorganism. As described above, the target data may include target spectral data and a target category. The acquisition unit 12 may read the target data from a predetermined storage device in response to an instruction signal from the user terminal 30. As another example, the acquisition unit 12 may receive the target data from an analysis device such as a laser Raman microscope in response to an instruction signal from the user terminal 30. Alternatively, the acquisition unit 12 may receive the target data transmitted from the user terminal 30.
[0044] In step S12, the first prediction unit 13 identifies one or more candidate countermeasures for the target microorganism using the countermeasure prediction model 41. In the present disclosure, this process is also referred to as "first identification process". The first prediction unit 13 inputs the target spectral data included in the target data into the countermeasure prediction model 41. The countermeasure prediction model 41 processes the countermeasure spectral data and predicts one or more candidate countermeasures effective for the target microorganism. The first prediction unit 13 acquires the prediction result.
[0045] In step S13, the second prediction unit 14 predicts the bacterial species of the target microorganism using the bacterial species prediction model 42, and identifies one or more candidate countermeasures corresponding to the bacterial species. In the present disclosure, this process is also referred to as "second identification process". The second prediction unit 14 inputs the target spectral data into the bacterial species prediction model 42. The bacterial species prediction model 42 processes the target spectral data to predict the bacterial species of the target microorganism. The second prediction unit 14 refers to the first countermeasure data 22a in the countermeasure database 22 and identifies one or more candidate countermeasures corresponding to the bacterial species indicated by the prediction result.
[0046] In step S14, the search unit 15 identifies one or more countermeasures corresponding to the target category. In the present disclosure, this process is also referred to as "third identification process". The search unit 15 refers to the second countermeasure data 22b in the countermeasure database 22 and identifies one or more candidate countermeasures corresponding to the target category.
[0047] The proposed system 10 executes at least one of steps S12 to S14, that is, at least one of the first identification process, the second identification process, and the third identification process. As an example, the acquisition unit 12 may acquire target data further including a designation of the identification process to be executed by the proposed system 10, and each of the first prediction unit 13, the second prediction unit 14, and the search unit 15 may determine whether to execute the process according to the designation.
[0048] In step S15, the filtering unit 16 performs filtering on one or more candidate countermeasures that have been identified to determine the recommended countermeasures. The one or more candidate countermeasures to be filtered include at least one of the candidate countermeasures predicted by the countermeasure prediction model 41, the candidate countermeasures corresponding to the bacterial species predicted by the bacterial species prediction model 42, and the one or more candidate countermeasures retrieved by the retrieval unit 15. Therefore, the one or more candidate countermeasures to be filtered may include the one or more candidate countermeasures (candidate countermeasures obtained by the first identification process or the second identification process) identified based on the prediction results of the prediction model 40 and the one or more candidate countermeasures retrieved (candidate countermeasures obtained by the third identification process). In one example, the filtering unit 16 determines the candidate countermeasures that are common among the candidate countermeasures obtained by the first identification process, the candidate countermeasures obtained by the second identification process, and the candidate countermeasures obtained by the third identification process as the recommended countermeasures. That is, the filtering unit 16 determines the candidate countermeasures identified in all of these three types of identification processes as the recommended countermeasures. When two of the three identification processes are specified by the target data, the filtering unit 16 determines the candidate countermeasures that are common between the two specified identification processes as the recommended countermeasures. When one of the three identification processes is specified by the target data, the filtering unit 16 determines the candidate countermeasures identified by the specified identification process as the recommended countermeasures as they are.
[0049] As an example, assume that the first identification process, the second identification process, and the third identification process are executed, and candidate countermeasures are identified as follows in each process. In this case, the filtering unit 16 determines sodium acetate, glycine, and low-temperature management as the recommended countermeasures. · First identification process: Sodium acetate, glycine, heat treatment, anaerobic packaging, low pH, and low-temperature management. · Second identification process: Sodium acetate, glycine, and low-temperature management. · Third identification process: Sodium acetate, glycine, heat treatment, anaerobic packaging, and low-temperature management.
[0050] In step S16, the output unit 17 outputs the processing result. The output unit 17 generates result data indicating at least recommended countermeasures. The output unit 17 may further generate result data indicating information regarding the predicted bacterial species. The output unit 17 transmits the generated result data to the user terminal 30. The user terminal 30 receives and displays the result data, whereby the user can obtain information such as recommended countermeasures.
[0051] FIG. 6 is a diagram showing an example of displaying the processing result. The screen 200 shown in this example shows Bacillus cereus, which is the predicted bacterial species of the target microorganism, information regarding the characteristics of this bacterial species, and two recommended countermeasures, countermeasure A and countermeasure B. Countermeasure A indicates the recommended countermeasure "sodium acetate", and countermeasure B indicates the recommended countermeasure "low temperature control".
[0052] [Modification Example] As described above, the technology according to the present disclosure has been described in detail based on various examples thereof. However, the present disclosure is not limited to the above examples. Various modifications are possible for the technology according to the present disclosure without departing from the gist thereof.
[0053] In the above example, the proposal system 10 includes the learning unit 11. However, a computer system different from the proposal system 10 may generate a prediction model, and the prediction model may be transplanted into the proposal system and used.
[0054] In the above example, the proposal system 10 includes the first prediction unit 13, the second prediction unit 14, and the search unit 15. However, the proposal system may not include a search unit. Alternatively, the proposal system may not include at least one of the first prediction unit and the second prediction unit. Therefore, the proposal system may determine any of the following candidate countermeasures as the recommended countermeasure. · Candidate countermeasures identified in all of the first specific process, the second specific process, and the third specific process. · Candidate countermeasures identified in both the first specific process and the second specific process. · Candidate countermeasures identified in both the first specific process and the third specific process. ·Candidate countermeasures identified in both the second specific process and the third specific process. ·Candidate countermeasures identified in the first specific process. ·Candidate countermeasures identified in the second specific process.
[0055] In any case, the proposal system inputs the target spectral data into a prediction model learned to execute predictions based on the spectral data of microorganisms, and obtains a prediction result regarding the target microorganism. This prediction result includes at least one of one or more candidate countermeasures predicted by the countermeasure prediction model and the bacterial species predicted by the bacterial species prediction model. Based on the prediction result, the proposal system determines at least one of the plurality of candidate countermeasures as the recommended countermeasure. For example, the proposal system may determine at least one of one or more candidate countermeasures indicated by the prediction result of the countermeasure prediction model as the recommended countermeasure. Alternatively, the proposal system may determine one or more candidate countermeasures corresponding to the bacterial species predicted by the bacterial species prediction model as the recommended countermeasure. Alternatively, the proposal system may determine the recommended countermeasure based on the prediction result by the prediction model and one or more candidate countermeasures retrieved from the countermeasure database.
[0056] Alternatively, the proposal system may include a search unit and may not include the first prediction unit and the second prediction unit. That is, the proposal system may not use a prediction model.
[0057] In the above example, the proposal system 10 is constructed as a server in a client-server system. As another example, the proposal system may be implemented on a stand-alone computer. Alternatively, the proposal system may be implemented on a user terminal accessible via a communication network to a predetermined database.
[0058] In the present disclosure, the expression "at least one processor executes a first process, executes a second process, ..., executes an nth process." or a corresponding expression is a concept that includes cases where the execution subject (i.e., the processor) changes midway among the n processes from the first process to the nth process. That is, this expression is a concept that includes both cases where all of the n processes are executed by the same processor and cases where the processor changes in an arbitrary manner among the n processes.
[0059] The processing procedures of the method executed by at least one processor are not limited to the above example. For example, some of the above-described steps may be omitted, or each step may be executed in a different order. Also, any two or more of the above-described steps may be combined, or a part of a step may be modified or deleted. Alternatively, other steps may be executed in addition to each of the above steps.
[0060] When comparing the magnitude relationship between two numerical values in a computer system or a computer, either of the two criteria of "greater than or equal to" and "greater than" may be used, and either of the two criteria of "less than or equal to" and "less than" may be used.
[0061] [Appendix] As can be understood from the various examples above, the present disclosure includes the aspects shown below. (Appendix 1) Comprising at least one processor, wherein the at least one processor, acquires target spectral data obtained by analyzing a target microorganism, inputs the target spectral data into a prediction model learned to execute a prediction based on spectral data of microorganisms, and obtains a prediction result regarding the target microorganism, based on the prediction result, determines at least one of a plurality of candidate countermeasures preset as countermeasures for the microorganism as a recommended countermeasure for the target microorganism, proposal system. (Appendix 2) The prediction model includes a countermeasure prediction model that is trained to receive the spectral data of the microorganism and predict one or more countermeasures against the microorganism, The at least one processor inputs the target spectral data into the countermeasure prediction model, obtains the prediction result indicating one or more of the candidate countermeasures predicted by the countermeasure prediction model, and determines at least one of the one or more candidate countermeasures indicated by the prediction result as the recommended countermeasure among the plurality of candidate countermeasures. The proposal system according to Supplementary Note 1. (Supplementary Note 3) The prediction model includes a strain prediction model that is trained to receive the spectral data of the microorganism and predict the strain of the microorganism, The at least one processor inputs the target spectral data into the strain prediction model, obtains the prediction result indicating the strain predicted by the strain prediction model, refers to a countermeasure database showing the correspondence between a plurality of the strains and the plurality of candidate countermeasures, and determines one or more candidate countermeasures corresponding to the strain indicated by the prediction result as the recommended countermeasure. The proposal system according to Supplementary Note 1 or 2. (Supplementary Note 4) The at least one processor further obtains a target category, which is a food and drink category corresponding to the target microorganism, refers to a countermeasure database showing the correspondence between a plurality of the food and drink categories and the plurality of candidate countermeasures, searches for one or more candidate countermeasures corresponding to the target category, and determines at least one of the plurality of candidate countermeasures as the recommended countermeasure based on the prediction result and the one or more candidate countermeasures retrieved. The proposal system according to any one of Supplementary Notes 1 to 3. (Supplementary Note 5) The at least one processor determines, as the recommended countermeasure, a common countermeasure among the one or more candidate countermeasures corresponding to the prediction result and the one or more candidate countermeasures retrieved. The proposal system according to Supplementary Note 4. (Supplementary Note 6) A proposal method executed by a proposal system including at least one processor, acquiring target spectral data obtained by analyzing a target microorganism; inputting the target spectral data into a prediction model learned to perform a prediction based on spectral data of microorganisms, and obtaining a prediction result regarding the target microorganism; determining, as a recommended countermeasure for the target microorganism, at least one of a plurality of candidate countermeasures preset as countermeasures for the microorganism based on the prediction result; A proposal method including the above. (Supplementary Note 7) acquiring target spectral data obtained by analyzing a target microorganism; inputting the target spectral data into a prediction model learned to perform a prediction based on spectral data of microorganisms, and obtaining a prediction result regarding the target microorganism; determining, as a recommended countermeasure for the target microorganism, at least one of a plurality of candidate countermeasures preset as countermeasures for the microorganism based on the prediction result; A proposal program that causes a computer to execute the above.
[0062] According to Supplementary Notes 1, 6, and 7, based on a prediction result obtained by inputting target spectral data of a target microorganism into a prediction model obtained by learning, a countermeasure for the target microorganism is determined from a plurality of candidate countermeasures. By introducing a prediction model, a countermeasure for a microorganism can be quickly proposed as long as spectral data of the microorganism is prepared. For example, a countermeasure can be quickly proposed for an unknown target microorganism.
[0063] According to Supplementary Note 2, since countermeasures are directly derived from the target spectrum data by the target prediction model obtained through learning, the configuration of the proposed system can be simplified accordingly.
[0064] According to Supplementary Note 3, the bacterial species is predicted from the target spectrum data by the bacterial species prediction model obtained through learning, and the countermeasures corresponding to the bacterial species are obtained from the database. By considering the bacterial species, the recommended countermeasures can be determined more accurately.
[0065] According to Supplementary Note 4, the recommended countermeasures are determined by considering not only the prediction result by the prediction model but also the candidate countermeasures corresponding to the food and beverage category corresponding to the target microorganism. By this method, the recommended countermeasures can be determined considering the characteristics of the food and beverage.
[0066] According to Supplementary Note 5, the candidate countermeasures specified in both the prediction result and the search based on the food and beverage category are determined as the recommended countermeasures. By this method, appropriate recommended countermeasures can be determined from the viewpoints of both microorganisms and food and beverages.
Explanation of Reference Signs
[0067] 10…Proposed system, 11…Learning unit, 12…Acquisition unit, 13…First prediction unit, 14…Second prediction unit, 15…Search unit, 16…Filtering unit, 17…Output unit, 21…Learning database, 22…Countermeasure database, 30…User terminal, 40…Prediction model, 41…Countermeasure prediction model, 42…Bacterial species prediction model, 110…Proposed program.
Claims
1. Comprising at least one processor, wherein the at least one processor acquires target spectrum data obtained by analyzing a target microorganism, inputs the target spectrum data into a prediction model learned to execute a prediction based on the spectrum data of microorganisms, and acquires a prediction result regarding the target microorganism, and based on the prediction result, determines at least one of a plurality of candidate countermeasures preset as countermeasures for the microorganism as a recommended countermeasure for the target microorganism. A proposal system.
2. The prediction model includes a countermeasure prediction model learned to receive the spectrum data of the microorganism and predict one or more countermeasures for the microorganism, wherein the at least one processor inputs the target spectrum data into the countermeasure prediction model, acquires the prediction result indicating one or more of the candidate countermeasures predicted by the countermeasure prediction model, and determines at least one of the one or more candidate countermeasures indicated by the prediction result among the plurality of candidate countermeasures as the recommended countermeasure. The proposal system according to claim 1.
3. The prediction model includes a strain prediction model learned to receive the spectrum data of the microorganism and predict the strain of the microorganism, wherein the at least one processor inputs the target spectrum data into the strain prediction model, acquires the prediction result indicating the strain predicted by the strain prediction model, and determines, with reference to a countermeasure database indicating the correspondence between the plurality of strains and the plurality of candidate countermeasures, one or more candidate countermeasures corresponding to the strain indicated by the prediction result as the recommended countermeasure. The proposal system according to claim 1 or 2.
4. The at least one processor further acquires a target category, which is a food and beverage category corresponding to the target microorganism, searches for one or more candidate countermeasures corresponding to the target category with reference to a countermeasure database indicating the correspondence between the plurality of food and beverage categories and the plurality of candidate countermeasures, and based on the prediction result and the one or more candidate countermeasures retrieved, determines at least one of the plurality of candidate countermeasures as the recommended countermeasure. The proposal system according to claim 1 or 2.
5. The at least one processor determines, as the recommended countermeasure, a countermeasure that is common between the one or more candidate countermeasures corresponding to the prediction result and the one or more candidate countermeasures retrieved. The proposal system according to claim 4. **Claim 6** A proposal method executed by a proposal system including at least one processor, comprising: obtaining target spectrum data obtained by analyzing a target microorganism; inputting the target spectrum data into a prediction model learned to perform a prediction based on spectrum data of microorganisms, and obtaining a prediction result regarding the target microorganism; determining, based on the prediction result, at least one of a plurality of candidate countermeasures preset as countermeasures for the microorganism as a recommended countermeasure for the target microorganism; The proposal method including the above steps. **Claim 7** obtaining target spectrum data obtained by analyzing a target microorganism; inputting the target spectrum data into a prediction model learned to perform a prediction based on spectrum data of microorganisms, and obtaining a prediction result regarding the target microorganism; determining, based on the prediction result, at least one of a plurality of candidate countermeasures preset as countermeasures for the microorganism as a recommended countermeasure for the target microorganism; A proposal program that causes a computer to execute the above steps.