JP7740943B2 - Information processing system and information processing method - Google Patents

Description

The present invention relates to information processing related to loss of items resulting from individual actions.

Patent Document 1 describes an information processing system that analyzes the relationship between lifestyle habits and physical condition and provides information for improving physical condition. Patent Document 2 describes a production and distribution management system that matches supply and demand based on product supply forecasts from multiple producers and order status from multiple consumers.

Patent No. 6904625 Japanese Patent Application Laid-Open No. 2002-024702

For example, when it comes to items with expiration dates, such as food, not all purchased items are necessarily consumed, and some may be discarded without being consumed. Therefore, there is a need for technology that reduces losses of items due to the actions of individuals, such as consumers, such as post-purchase losses of food. The technology described in Patent Document 1 is a technology for improving physical condition, and is not related to the loss of items. The technology described in Patent Document 2 is a technology for matching supply and demand between producers' supply forecasts and consumers' orders, and cannot address losses of items due to the actions of individuals, such as post-purchase food waste.

The object of the present invention is to provide an information processing system that reduces the loss of items based on individual behavior.

An information processing system according to one embodiment of the present invention includes:
a behavior prediction means for calculating a behavior probability that an individual will perform a predetermined behavior based on the behavior history of the individual;
and loss determination means for calculating the probability of loss of a predetermined item due to an individual's action based on the action probability, determining whether or not the action probability needs to be changed based on the action probability and the loss probability, and outputting an instruction to change the action probability if it is determined that the action probability needs to be changed.

An information processing method according to one aspect of the present invention includes:
The behavior prediction device calculates a behavior probability that the individual will perform a predetermined behavior based on the individual's behavior history;
A loss determination device calculates the loss probability of a predetermined item accompanying an individual's behavior based on the behavior probability, determines whether or not the behavior probability needs to be changed based on the behavior probability and the loss probability, and outputs an instruction to change the behavior probability when it is determined that the behavior probability needs to be changed.

This invention has the effect of reducing the loss of items due to individual actions.

1 is a block diagram showing an example of a configuration of an information processing system according to a first embodiment of the present invention. FIG. 10 is a block diagram illustrating an example of a configuration of an information processing system according to a second embodiment. FIG. 10 is a diagram showing an example of a food purchase history. FIG. 10 is a diagram illustrating an example of a notification output by a behavior prediction node to a consumer. FIG. 10 is a diagram showing an example of food waste history. 10A and 10B are diagrams illustrating examples of instructions to change a food waste determination node and changes to food purchase probabilities in a behavior prediction node. FIG. 1 is a diagram showing an example of the relationship between a consumer's food purchase probability and an intermediary's food waste rate. FIG. 10 is a diagram illustrating an example of an instruction for a production quantity to a production management node. This is a diagram showing changes in food waste rates between consumers and intermediaries. 3 is a block diagram showing an example of a configuration of an information processing system including the information processing system of FIG. 2. FIG. FIG. 2 is a block diagram showing a configuration of an information processing device as an example of a hardware configuration of a node.

Next, embodiments of the present invention will be described with reference to the drawings. Each drawing is intended to explain an embodiment of the present invention, but does not limit the embodiments of the present invention. Furthermore, each embodiment can be combined as appropriate. Furthermore, the direction of the arrows in the drawings is an example, and does not limit the direction of signals between blocks.

First Embodiment
An information processing system 11 according to a first embodiment will be described with reference to the drawings. FIG. 1 is a block diagram showing an example of the configuration of the information processing system 11 according to the first embodiment. The information processing system 11 includes a behavior prediction device 111 and a loss determination device 121. The behavior prediction device 111 calculates a behavior probability that an individual will perform a predetermined behavior based on the behavior history of the individual. The loss determination device 121 calculates a loss probability of a predetermined item associated with the individual's behavior based on the behavior probability, determines whether or not the behavior probability needs to be changed based on the behavior probability and the loss probability, and outputs an instruction to change the behavior probability when it is determined that the behavior probability needs to be changed.

The information processing system 11 configured in this manner calculates the probability of loss of an item based on the behavioral probability, and outputs an instruction to change the behavioral probability based on the calculated loss probability. The instruction to change the behavioral probability is a change to the predicted behavior of an individual's item. Therefore, the behavior prediction device 111 or loss determination device 121 may instruct devices, etc. included in the item's value chain to perform an action to reduce the loss of the item based on this change instruction. In this case, the devices, etc. included in the item's value chain can perform an action to reduce the loss of the item based on this instruction.

Alternatively, the information processing system 11 may output an instruction to change the action probability, and each device included in the value chain of the product may operate based on the instruction to change the action probability output by the information processing system 11. For example, a device managing the behavior of individuals may output an instruction to change their behavior to an individual based on the instruction to change the action probability output by the information processing system 11. For example, in the case of an instruction to lower the action probability, the device managing the behavior of individuals may output a message to the individual to refrain from taking action toward the product. Conversely, in the case of an instruction to increase the action probability, the device managing the individual may output a message to the individual encouraging the action. Alternatively, a device managing the sales of products may change the price or distribution quantity of the product in response to an instruction to change the action probability. Alternatively, a device managing the production of products may change the production quantity of the product in response to an instruction to change the action probability.

The information processing system 11 may also use behaviors other than those related to the loss of property as the target of judgment for individual behavior. For example, the information processing system 11 may use the behavior of an individual during a disaster as the target individual behavior to be judged. For example, the information processing system 11 may calculate the probability of evacuation behavior as the individual behavior during a disaster, and determine the need for disaster-related actions based on the calculated probability of evacuation behavior. For example, the information processing system 11 may determine the need for disaster notification and evacuation support for people who have difficulty evacuating, such as the elderly, based on the probability of evacuation behavior. In this way, the information processing system 11 may operate not only for property loss, but also for other events such as disasters. In other words, the information processing system 11 outputs appropriate information at the appropriate time regarding events such as property loss based on the coordination of the operations of the behavior prediction device 111 and the loss determination device 121. As a result, the information processing system 11 realizes part of a fair and equal social solution based on the operations between the components.

Second Embodiment
The second embodiment will be described below with reference to the drawings. In the second embodiment, an example will be described in which, in a system including multiple nodes that perform operations, the functions of the information processing system 11 according to the first embodiment are used as basic functions, and at least one of the nodes performs the functions of the first embodiment. The nodes that perform operations may be individual computer devices connected to a network or computers virtually implemented in cloud computing. However, the method of implementing the nodes is not limited to these. For example, the node may be a system combining multiple computer devices connected to a network or a system combining multiple computers virtually implemented in cloud computing. Alternatively, a single computer device or virtual computer may perform the functions of multiple nodes. As such, a node may be a physical device or a combination of software and hardware. Therefore, the node in the following description may be a device, a system, a component included in a single device, or a means for implementing a function. The hardware configuration of a computer device when a node is implemented using a computer device will be described later.

The nodes may be connected using dedicated lines, public telephone lines, or general-purpose communication networks such as the Internet. If the nodes are virtual computers, the connections may be virtual networks. The connection structure may be a non-hierarchical structure such as a ring, mesh, or star structure, or a hierarchical structure such as a neural structure. However, the connection method and structure are not limited to the above.

In the following description, at least some nodes may use artificial intelligence (AI) in their respective judgments. When using AI, the node specifically uses a trained model obtained using machine learning. For example, a node using AI may use a trained model obtained using heterogeneous mixture learning, an example of machine learning. Heterogeneous mixture learning is a type of machine learning that automatically discovers specific patterns from the correlations between data mixed in a wide variety of data and switches the rules it references depending on the data being analyzed. However, nodes using AI are not limited to heterogeneous mixture learning, and may use other learning methods as AI, such as support vector machines, naive Bayes, or deep learning using neural networks. Furthermore, a node using AI may re-train a trained model used for judgment using the results of judgments made by other nodes and information from nodes not shown. In this way, nodes using AI may coordinate operations such as judgments.

Furthermore, the second embodiment uses food purchases as an example of behavior and food waste as an example of loss. In this embodiment, food waste refers to "all food that is discarded despite being edible." In other words, in the following description, food waste includes "food loss" and "food waste" as defined in "The State of Food and Agriculture, 2019 (SOFA)" published by the Food and Agriculture Organization of the United Nations. In SOFA, food loss refers to the reduction in food quantity or quality caused by the decisions or actions of food suppliers within the chain, excluding retailers, food service providers, and consumers. Furthermore, food waste refers to the reduction in food quantity or quality caused by the decisions or actions of retailers, food service providers, and consumers. In other words, this embodiment targets the reduction of food waste throughout the entire food value chain.

In addition, the following explanation will focus on food waste at the consumer level within the food value chain. This is because targeting consumers, who are the most numerous component in the overall structure of the food value chain, is likely to improve fairness throughout the food value chain and reduce information disparities throughout the entire food value chain. However, this does not mean that the scope of this embodiment is limited to food waste at the consumer level. As explained below, this embodiment may also perform operations to reduce food waste in other components of the food value chain.

The configuration of the information processing system 10 according to the second embodiment will be described with reference to the drawings. Figure 2 is a block diagram showing an example of the configuration of the information processing system 10 according to the second embodiment. The information processing system 10 includes a behavior prediction node 110, a food loss determination node 120, a product management node 130, and a production management node 140. The number of nodes shown in Figure 2 is an example. The information processing system 10 may include multiple nodes as at least some of the nodes. For example, the information processing system 10 may include multiple behavior prediction nodes 110 corresponding to each target consumer. Alternatively, the information processing system 10 may include multiple product management nodes 130 corresponding to each seller of the target food. Alternatively, the information processing system 10 may include multiple production management nodes 140 corresponding to each producer of the target food. In this way, the information processing system 10 may operate by including a node corresponding to each of multiple targets, such as consumers.

Alternatively, each node may operate on multiple targets, rather than a single target. For example, the behavior prediction node 110 may predict the behavior of multiple consumers. When operating on multiple targets, each node may execute the operations for each processing target in tandem, or may execute at least some of the operations in parallel. However, in the following description, to avoid complexity, the information processing system 10 includes only one node as each node. Furthermore, in the following description, as an example, each node operates on one food item and one consumer. Furthermore, each node shown in FIG. 2 uses a trained model obtained using AI in its respective judgments. Note that each node may re-execute machine learning using the results of judgments at other nodes and information from nodes not shown in the figure to improve the performance of the trained model used for judgment.

The information processing system 10 begins operation when conditions set by a system user are satisfied. For example, the information processing system 10 may begin operation upon acquiring information on the target food and consumer. For example, the information processing system 10 may begin operation when acquiring information on the food and consumer from a node that determines the target food and consumer. The node that determines the target food and consumer is, for example, a node that determines the food and consumer using image recognition using images captured by a camera near the food. Alternatively, the information processing system 10 may acquire the storage location of the target food and consumer information from a user, etc., and operate using the stored information on the food and consumer. In this way, the information processing system 10 may begin operation in response to any operation start condition. Therefore, in the following explanation, it is assumed that the information processing system 10 has already acquired information on the target food and consumer. For example, the behavior prediction node 110 has already acquired information on the target food and consumer in the following explanation. Therefore, for example, the acquisition of personal information, etc. by the behavior prediction node 110, described later, is the acquisition of personal information for consumers whose information has already been acquired. The same applies to food.

[Behavior prediction node 110]
The behavior prediction node 110 stores the consumer's food purchase history as the individual's behavior history, and calculates the consumer's food purchase probability as the individual's behavior probability. In other words, the behavior prediction node 110 is an example of the behavior prediction device 111. Furthermore, the behavior prediction node 110 stores a food purchase probability calculation model that calculates the consumer's food purchase probability, and calculates the consumer's food purchase probability using the food purchase probability calculation model. In detail, the behavior prediction node 110 operates as follows. The behavior prediction node 110 previously stores at least food information, the consumer's personal information, the consumer's behavior data, and the food purchase probability as the consumer's past food purchase history, in association with each other.

Food information is information that includes at least one of a food identifier and food characteristics. The food identifier is, for example, the food name or the food product code. The food characteristics are, for example, fresh food, processed food, or beverage. Alternatively, the food characteristics are, for example, new product, seasonal limited edition, or regular. However, food information is not limited to the above and may be other information depending on the food in question. Personal information is information that identifies the consumer. For example, personal information may include at least one of occupation, age, gender, and physique. However, personal information is not limited to the above and may also include biometric information of the consumer or information other than the above, such as current location. Biometric information is, for example, information that includes at least one of body temperature, heart rate, and sweat rate. However, biometric information is not limited to the above and may also include information other than the above, such as the name of a disease. Behavioral data is data related to the consumer's behavior. For example, behavioral data may include at least one of walking, running, sitting, traveling by vehicle, commuting to work, and returning home. However, behavioral data is not limited to the above and may include information other than the above, such as the consumer's range of daily activities.

The behavior prediction node 110 may store other information related to the consumer and food in addition to the above as food purchase history. Alternatively, the behavior prediction node 110 may store other information related to food purchases in addition to the above as food purchase history. For example, weather may affect food purchases. Therefore, the behavior prediction node 110 may store weather information as food purchase history.

Figure 3 is a diagram showing an example of a food purchase history. The food purchase history in Figure 3 includes occupation, which is an example of personal information, behavioral data, food information, and food purchase probability. The food purchase history in Figure 3 also includes weather information and biometric information. For example, the first line in Figure 3 indicates that "a high school student with a normal body temperature returning from club activities will purchase a newly released food product with an 80% probability if the weather is sunny and the temperature is 25°C."

The behavior prediction node 110 then acquires personal information and behavioral data of the consumer to be judged. For example, the behavior prediction node 110 acquires the consumer's personal information from a node that stores the consumer's personal information, and acquires the consumer's behavioral data from a node that monitors the consumer's behavior. For example, the behavior prediction node 110 may acquire the consumer's behavioral data from a node that manages sensors that measure the consumer's behavior in real time. Alternatively, the behavior prediction node 110 may acquire at least a portion of the consumer's behavioral history from a device that stores the consumer's behavioral history, in addition to or instead of real-time information. However, the behavior prediction node 110 is not limited to acquiring personal information and behavioral data from the above nodes, and may also acquire personal information and behavioral data from nodes other than the above nodes. The behavior prediction node 110 may acquire other information in addition to the consumer's personal information and behavioral data. For example, the behavior prediction node 110 may acquire the consumer's biometric information from a node that measures the consumer's biometric information. Alternatively, the behavior prediction node 110 may acquire meteorological information from a node that provides meteorological information such as weather and temperature.

The behavior prediction node 110 then inputs the stored food purchase history, the acquired food information, personal information, and behavioral data into a food purchase probability calculation model, and obtains the consumer's food purchase probability from the food purchase probability calculation model. The behavior prediction node 110 may input other information, such as weather information, in addition to the food purchase history, food information, personal information, and behavioral data, into the food purchase probability calculation model to obtain the consumer's food purchase probability. The behavior prediction node 110 then outputs the acquired food information and personal information, the food purchase probability obtained from the food purchase probability calculation model, and the food purchase history to the food waste determination node 120. The behavior prediction node 110 may output other information, such as behavioral data, in addition to the food information, personal information, food purchase probability, and food purchase history.

When the behavior prediction node 110 receives an instruction to change the food purchase probability from the food waste determination node 120, it changes the food purchase probability in the food purchase history based on the instruction to change. Furthermore, the behavior prediction node 110 may output a food purchase instruction related to the purchase to the consumer based on the instruction to change the food purchase probability. For example, if the instruction to change the food purchase probability is to lower the food purchase probability, the behavior prediction node 110 should restrict the consumer's purchases. Therefore, in this case, the behavior prediction node 110 may output an instruction to the consumer to prevent the purchase of the food. For example, the behavior prediction node 110 may output an instruction to prevent the purchase of the food to a node that measures the consumer's biometric information. Alternatively, the behavior prediction node 110 may output a notification to a terminal device carried by the consumer to suspend the purchase of the food.

On the other hand, if the instruction to change the food purchase probability is an instruction to increase the food purchase probability, the behavior prediction node 110 should encourage the consumer to purchase. Therefore, in this case, for example, the behavior prediction node 110 may output an instruction to purchase food to a node that manages the consumer's health. Alternatively, the behavior prediction node 110 may output a notification to a terminal device carried by the consumer to encourage the purchase of food. For example, the behavior prediction node 110 may output a food "recommendation notification" to the consumer. Figure 4 is a diagram showing an example of a notification output by the behavior prediction node 110 to the consumer. For example, in the first line of Figure 4, the behavior prediction node 110 has received an instruction to change the food purchase probability to decrease the food purchase probability, so it decreases the food purchase probability from 70% to 60%. The behavior prediction node 110 then outputs a "purchase pending" notification to prevent the consumer from making a purchase.

The behavior prediction node 110 may output instructions to other nodes related to food. For example, if the instruction to change the food purchase probability is to lower the food purchase probability, it is desirable to curb food purchases. Therefore, the behavior prediction node 110 may output an instruction to the product management node 130 to increase the price of the food. Conversely, if the instruction to change the food purchase probability is to increase the food purchase probability, it is desirable to encourage food purchases. Therefore, for example, the behavior prediction node 110 may output an instruction to the product management node 130 to lower the price of the food.

[Food loss determination node 120]
The food loss determination node 120 calculates the food loss rate as the probability of losing an item. Furthermore, the food loss determination node 120 determines whether or not a change to the food purchase probability is necessary as a change to the behavioral probability. Furthermore, the food loss determination node 120 outputs an instruction to change the food purchase probability as an instruction to change the behavioral probability. In other words, the food loss determination node 120 is an example of a loss determination device 121. Furthermore, the food loss determination node 120 stores a food loss rate calculation model for calculating the food loss rate, and calculates the food loss rate using the food loss rate calculation model. In detail, the food loss determination node 120 operates as follows.

The food loss determination node 120 stores in advance, as a past food loss history, at least food information, consumer personal information, and food loss rates, in association with each other. For example, the food loss determination node 120 stores food loss rates corresponding to foods and consumers as food loss history. The food loss determination node 120 may store one food loss rate for each food. Alternatively, the food loss determination node 120 may store multiple food loss rates corresponding to multiple parts of the value chain. For example, the food loss determination node 120 may store food loss rates for producers, intermediaries, and consumers that make up the value chain. In this embodiment, producers include those in primary industries such as agriculture, forestry, or fishing, as well as those in secondary industries such as food manufacturers. Intermediaries include wholesalers and retailers, as well as transporters and warehouse operators, who are part of the value chain excluding producers and consumers.

Figure 5 shows an example of a food waste history. The food waste history in Figure 5 includes occupation, which is an example of personal information about the consumer, food information, and the food waste rates for the consumer, intermediary, and producer. For example, the first line shows that the food waste rate for a new product is 10% for the consumer (i.e., high school student) after purchase, 25% for the intermediary, and 5% for the producer.

The food waste determination node 120 then inputs the food information, personal information, food purchase probability, food purchase history, and stored food waste history obtained from the behavior prediction node 110 into a food waste rate calculation model, and obtains the food waste rate from the food waste rate calculation model. The food waste determination node 120 may use other information, such as weather information, in addition to food information. For example, the food waste determination node 120 may use weather information for the location where the food is sold. When weather information is used, the food waste determination node 120 can obtain the weather information from a specified node, just like the behavior prediction node 110.

The food loss determination node 120 then determines whether or not a change to the food purchase probability is necessary based on the food loss rate obtained from the food loss rate calculation model and the food loss history. Specifically, the food loss determination node 120 compares the food loss rate obtained from the food loss rate calculation model with the food loss rate included in the food loss history, and if the values differ, determines that a change to the food purchase probability is necessary. On the other hand, if the values are the same, the food loss determination node 120 determines that a change to the food purchase probability is unnecessary. If the food loss rate corresponds to each component of the value chain, such as consumers, the food loss determination node 120 uses the consumer's food loss rate to determine whether or not a change to the food purchase probability is necessary. Note that, taking into account determination error, the food loss determination node 120 may determine that a change to the food purchase probability is unnecessary if the difference in food loss rate is within a specified range.

If it is determined that a change in the food purchase probability is necessary, the food waste determination node 120 outputs a change instruction to the behavior prediction node 110 to change the food purchase probability. For example, if the food waste rate of the consumer determined by the food waste rate calculation model is higher than the food waste rate of the consumer in the food waste history, the food waste determination node 120 outputs a change instruction to the behavior prediction node 110 to lower the food purchase probability. Conversely, if the food waste rate of the consumer determined by the food waste rate calculation model is lower than the food waste rate of the consumer in the food waste history, the food waste determination node 120 outputs a change instruction to the behavior prediction node 110 to increase the food purchase probability.

Figure 6 is a diagram showing an example of a change instruction from the food waste determination node 120 and a change to the food purchase probability in the behavior prediction node 110. The upper diagram in Figure 6 is a diagram showing an example of a determination instruction output by the food waste determination node 120. In addition to the information in Figure 5, the upper diagram in Figure 6 includes the consumer's food waste rate determined by the food waste rate calculation model and a change instruction to the food purchase probability output by the food waste determination node 120 to the behavior prediction node 110. The lower diagram in Figure 6 is a diagram showing an example of a change to the food purchase probability in the behavior prediction node 110.

For example, in the first row of Figure 6, the consumer's food waste rate determined by the food waste rate calculation model (15%) is higher than the consumer's food waste rate in the food waste history (10%). Therefore, the food waste determination node 120 outputs a change instruction to the behavior prediction node 110 to "reduce the food purchase probability." In this case, the behavior prediction node 110 changes the food purchase history from 70% to 60% in response to the change instruction, as shown in the first row of the bottom diagram in Figure 6. Alternatively, in the second row of the top diagram in Figure 6, the consumer's food waste rate determined by the food waste rate calculation model (5%) is lower than the consumer's food waste rate in the food waste history (10%). Therefore, the food waste determination node 120 outputs a change instruction to the behavior prediction node 110 to "increase the food purchase probability." In this case, the behavior prediction node 110 changes the food purchase history from 70% to 80% in response to the instruction.

The food loss determination node 120 may output instructions to a node other than the behavior prediction node 110. For example, if the food loss rate of a consumer determined by the food loss rate calculation model is higher than the food loss rate of the consumer in the food loss history, the food loss determination node 120 may output a change instruction to the product management node 130 to reduce the distribution quantity of food. Alternatively, in this case, the food loss determination node 120 may output a change instruction to the production management node 140 to reduce the production quantity of food. Conversely, if the food loss rate of a consumer determined by the food loss rate calculation model is lower than the food loss rate of the consumer in the food loss history, the food loss determination node 120 may output a change instruction to the product management node 130 to increase the distribution quantity of food. Alternatively, in this case, the food loss determination node 120 may output a change instruction to the production management node 140 to increase the production quantity of food.

Note that if a consumer's food purchase probability changes, the quantity of food sold at a store will change, etc. Therefore, changing the food purchase probability may change the food waste rate at intermediaries. For example, lowering the food purchase probability in order to reduce the consumer's food waste rate may increase the food waste rate at intermediaries, including stores. Furthermore, a change in the amount of food distributed may change the food waste rate at producers. In this way, food waste may be correlated between different components of the value chain. Therefore, the food waste determination node 120 may take into account the correlation between the components included in the value chain. For example, the food waste determination node 120 may determine the consumer's food purchase probability to be changed using the correlation between the consumer's food waste rate and the food waste rate at intermediaries, in addition to the calculated consumer's food waste rate. In this case, the food waste determination node 120 may output the determined food purchase probability to the behavior prediction node 110 in addition to issuing an instruction to change the food purchase probability. In this case, the behavior prediction node 110 simply stores the acquired food purchase probability in the food purchase history.

Next, we will explain an example of a relationship between a consumer's food purchase probability and a middleman's food waste rate. Figure 7 shows an example of the relationship between a consumer's food purchase probability and a middleman's food waste rate. In Figure 7, the horizontal axis represents the consumer's food purchase probability. The vertical axis represents the middleman's food waste rate. The solid line represents the relationship between a consumer's food purchase probability and a middleman's food waste rate. For example, if a consumer's food purchase probability is 20%, the middleman's food waste rate is 58%. The dashed line represents the threshold for the food waste rate at the middleman. In Figure 7, the threshold is 35%. For example, suppose the behavior prediction node 110 calculates a consumer's food purchase probability to be 80%. Then, the food waste determination node 120 determines that it is desirable to lower the consumer's food purchase probability to 20% in order to reduce food waste. However, if the consumer's food purchase probability is lowered to 20%, the middleman's food waste rate will be 58%, exceeding the threshold of 35%. The food purchase probability required to keep the food waste rate of intermediaries to the threshold of 35% is 60%. Therefore, the food waste determination node 120 outputs the consumer food purchase probability of 60% to the behavior prediction node 110 in order to keep the food waste rate of intermediaries to the threshold of 35%.

Reducing food production volume will reduce the food waste rate at intermediaries. Conversely, increasing food production volume will increase the food waste rate at intermediaries. Therefore, the food waste determination node 120 may output a production volume change to the production management node 140 in order to reduce the food waste rate. An example of a production volume change instruction will be explained with reference to the drawings. Figure 8 is a diagram showing an example of a production volume change instruction sent to the production management node 140. Figure 9 is a diagram showing changes in food waste rates between consumers and intermediaries. In the following explanation, the upper limit for food waste rates at intermediaries is assumed to be 50%.

In Figures 8 and 9, the consumer's food waste rate decreases from 70% to 20% as the food purchase probability changes. Furthermore, as the food purchase probability changes, the food waste rate of intermediary A in the first row of Figure 8 increases from 30% to 35%. The diagram on the left side of Figure 9 shows the change in food waste rate between the consumer and intermediary A before and after the change in food purchase probability. In this case, the increased food waste rate for intermediary A does not exceed the upper limit. Therefore, the food waste determination node 120 does not output an instruction to the production management node 140. Meanwhile, the food waste rate of intermediary B in the second row increases from 30% to 60%. The diagram on the right side of Figure 9 shows the change in food waste rate between the consumer and intermediary B. In this case, the food waste rate for intermediary B after the change in food purchase probability exceeds the upper limit. Therefore, the food waste determination node 120 outputs an instruction to the production management node 140 to reduce the production quantity in order to reduce the intermediary's food waste rate.

[Product Management Node 130]
The product management node 130 manages food products sold to consumers in stores. Alternatively, the product management node 130 manages food products sold through mail order sales via a network. To this end, the product management node 130 stores a product management model that manages food prices and inventory quantities, and uses the product management model to manage food prices, as well as the quantity and timing of food products delivered to stores. For example, the product management node 130 changes the price of a food product based on a food price change instruction received from the behavior prediction node 110. Alternatively, the product management node 130 changes the delivery quantity and timing of food products based on a food distribution quantity change instruction received from the food loss determination node 120. For example, the product management node 130 may determine the delivery quantity and timing of food products using the current food inventory quantity obtained from a node that manages the sales and inventory quantities of products in a store or commercial area. For example, the product management node 130 inputs the product price changed in response to the price change instruction and the current food inventory quantity into the product management model, and obtains the appropriate inventory quantity from the product management model. The product management node 130 may output data on the food items whose prices have been changed to the behavior prediction node 110. In this case, the behavior prediction node 110 may calculate the food purchase probability using the acquired food item data.

[Production management node 140]
The production management node 140 manages food production. To this end, the production management node 140 stores a production quantity management model that manages production quantities and uses the production quantity management model to manage food production quantities. For example, the production management node 140 inputs into the production quantity management model a food production quantity change instruction obtained from the food loss determination node 120 and the current production status obtained from a node that monitors food production, to obtain an appropriate production quantity. The production management node 140 may use other information in addition to the above information. For example, the production management node 140 may use the inventory quantity of food stored by an intermediary such as a warehouse operator. For example, the production management node 140 may input the changed production quantity and the current inventory quantity of the warehouse operator into the production quantity management model, obtain the future food production quantity from the production quantity management model, and produce food based on the obtained production quantity.

[Information Processing System 10]
As described above, the information processing system 10 uses food purchase probability as an example of behavior probability and food waste rate as an example of loss probability, and operates with food consumers as individuals as its targets. In other words, the information processing system 10 executes coordinated operations among the nodes included in the information processing system 10, including the food waste determination node 120, based on the food purchase probability predicted by the behavior prediction node 110. The information processing system 10 then reduces food waste as a result of coordinated operations among multiple nodes. In other words, the information processing system 10 calculates the food purchase probability of a consumer predicted by the behavior prediction node 110 as an individual's behavior, and reduces food waste as a material loss based on the food purchase probability. In this way, the information processing system 10 according to the second embodiment reduces food waste, an example of a material loss. This is because the behavior prediction node 110 and the food waste determination node 120 perform operations similar to those of the behavior prediction device 111 and loss determination device 121 of the first embodiment, targeting food waste. In other words, the information processing system 10 is an example of the information processing system 11.

More specifically, the information processing system 10 is as follows. The information processing system 10 includes a behavior prediction node 110, which is an example of a behavior prediction device 111, and a food waste determination node 120, which is an example of a loss determination device 121. The behavior prediction node 110 includes a food purchase history including at least food information, personal information of the consumer, behavioral data of the consumer, and a food purchase probability, which is an example of a behavioral probability. The behavior prediction node 110 then acquires the food information and the personal information and behavioral data of the consumer. The behavior prediction node 110 then calculates the food purchase probability based on the food purchase history and the acquired food information, personal information, and behavioral data. The behavior prediction node 110 then outputs the acquired food information and personal information, the calculated food purchase probability, and the food purchase history to the food waste determination node 120. The behavior prediction node 110 then changes the food purchase probability in the food purchase history based on an instruction to change the food purchase probability acquired from the food waste determination node 120. The behavior prediction node 110 then outputs information related to food purchases to the consumer based on the change instruction. The food loss determination node 120 includes a food loss history that includes at least food information, personal information, and a food loss rate, which is an example of a loss probability. The food loss determination node 120 also acquires the food information, personal information, food purchase probability, and food purchase history from the behavior prediction node 110. The food loss determination node 120 then calculates the food loss rate based on the food loss history and the acquired food information, personal information, food purchase probability, and food purchase history. The food loss determination node 120 then determines whether or not to change the food purchase probability based on the calculated food loss rate. If the food loss determination node 120 determines that a change to the food purchase probability is necessary, it outputs an instruction to change the food purchase probability to the behavior prediction node 110. Based on this configuration, the information processing system 10 reduces food loss.

The information processing system 10 further includes a product management node 130 that manages food products sold to consumers. The behavior prediction node 110 then outputs a price change instruction to the product management node 130 based on the instruction to change the food product purchase probability. The product management node 130 then changes the price of the food product based on the price change instruction obtained from the behavior prediction node 110. Based on this configuration, the information processing system 10 can change the price of a product to reduce food waste. The food waste determination node 120 may then instruct the product management node 130 to change the distribution quantity of the food product based on the food waste rate. The product management node 130 may then change the transportation quantity of the food product based on the instruction to change the distribution quantity of the food product obtained from the food waste determination node 120. In this case, the information processing system 10 can further reduce food waste.

The information processing system 10 further includes a production management node 140 that manages food production. The food loss determination node 120 then outputs food production change instructions to the production management node 140 based on the food loss rate. The production management node 140 then changes the food production quantity based on the food production change instructions obtained from the food loss determination node 120. As a result, the information processing system 10 can further reduce food loss.

Note that the nodes included in FIG. 2 may be directly connected to relevant parties such as consumers, or may be indirectly connected via one or more nodes. For example, the behavior prediction node 110 may output information related to food purchases to a terminal device carried by the consumer, or may output information related to food purchases to a node that monitors the consumer's behavior, including food purchases. In this way, the information processing system 10 may operate in association with other nodes in addition to the nodes included in FIG. 2.

[Information Processing System 12]
FIG. 10 is a block diagram showing an example of the configuration of an information processing system 12 including the information processing system 10 of FIG. 2 . In addition to the nodes described above, the information processing system 12 includes a personal information node 210, a behavior monitoring node 220, a biological information measurement node 230, and a product data node 240. Furthermore, the information processing system 12 includes an agriculture management node 251, a fisheries management node 252, and a food production management node 253 as nodes for managing food production. Furthermore, the information processing system 12 includes a crop observation node 254, a fish catch observation node 255, and a food production monitoring node 256 as nodes for observing or monitoring food production. Furthermore, the information processing system 12 includes a weather information node 260 and a logistics management node 270. The information processing system 12 can use any connection method, such as a neural network, to connect the nodes. The number of nodes shown in FIG. 10 is merely an example. The information processing system 12 may include multiple nodes as at least some of the nodes. 10, at least some of the nodes may be nodes that use AI. In FIG. 10, nodes that use AI in the information processing system 12 are marked with "(AI)."

The personal information node 210 is a node that stores the consumer's personal information and outputs the personal information to the behavior prediction node 110. For example, the personal information node 210 may be a system or device of a business that acquires personal information from consumers and manages their assets and liabilities, such as a bank or credit company. Alternatively, the personal information node 210 may be a system or device of a business that acquires personal information from consumers and provides specific services to them, such as a music distribution site. However, the personal information node 210 is not limited to the above and may be other systems or devices. The behavior monitoring node 220 is a node that monitors the consumer's behavior and outputs the consumer's behavior data to the behavior prediction node 110. For example, the behavior monitoring node 220 may be a system or device that acquires information in real time from a sensor that detects at least one of the consumer's position, speed, and acceleration. Alternatively, the behavior monitoring node 220 may be a system or device that recognizes people in images acquired from surveillance cameras in stores, roads, etc. Note that the behavior monitoring node 220 may be a system or device that includes the above sensors. However, the behavior monitoring node 220 is not limited to the above and may be other systems or devices. The biological information measurement node 230 is a node that acquires the consumer's biological information and outputs the biological information to the behavior prediction node 110. For example, the biological information measurement node 230 may be a system or device that acquires information from a sensor that detects the consumer's body temperature, heart rate, or amount of sweat. Alternatively, the biological information measurement node 230 may be a system or device that manages a hospital, a sports gym, or the like. However, the biological information measurement node 230 is not limited to the above and may be another system or device.

The product data node 240 is a node that monitors the sales and inventory quantities of food in stores or business areas, such as retail stores and food service businesses, and outputs food inventory quantities and other information to the product management node 130. For example, the product data node 240 may be a server managed by the store, an inventory management server in the store's warehouse, or a point of sale (POS) server for the store. However, the product data node 240 is not limited to the above and may be another system or device.

The crop observation node 254 observes the state of the crops that farmers are cultivating to become food. For example, the crop observation node 254 is a node that observes the state of farms using drones equipped with cameras. The catch quantity observation node 255 observes the catch quantity of seafood that fishermen land to become food. For example, the catch quantity observation node is a node that obtains the landing status at the fishery market. The food production monitoring node 256 monitors the production status of food at food manufacturers. For example, the food production monitoring node 256 is a server that monitors the production line of a food factory. However, these nodes are not limited to the above.

The agricultural management node 251 acquires the status of agricultural crops from the agricultural crop observation node 254, determines the production quantity and possible shipping date of agricultural crops to be used as food, and outputs this information to the production management node 140. For example, the agricultural management node 251 is a device at an agricultural cooperative or a device at an agricultural crop wholesaler. The fishery management node 252 acquires the catch quantity of seafood from the catch quantity observation node 255 and outputs information about seafood to be used as food to the production management node 140. For example, the fishery management node 252 is a device at a fish market or a device at a seafood wholesaler. The food production management node 253 outputs the possible shipping date and possible shipping quantity of produced food to the production management node 140 based on the production status of food acquired from the food production monitoring node 256. For example, the food production management node 253 is a production management server at a food manufacturer. However, these nodes are not limited to the above. The agriculture management node 251, fisheries management node 252, and food production management node 253 in Figure 10 use AI. In other words, these nodes use trained models in their respective judgments. The production management node 140 manages food production based on information from each node.

The weather information node 260 is a node that outputs weather information to at least one of the behavior prediction node 110 and the food waste assessment node 120. The weather information node 260 may store a weather forecast model that predicts weather at a specified location and time. The weather information node 260 may then use the weather forecast model to predict weather at a location and time specified by the behavior prediction node 110 or the food waste assessment node 120, and output the predicted weather as weather information. For example, the weather information node 260 inputs information obtained from a weather information provider and the specified location and time into the weather forecast model, obtains weather information from the weather forecast model, and outputs the obtained weather information.

The logistics management node 270 manages the transportation of food. The logistics management node 270 may store a logistics management model that manages the transportation quantity and transportation time of food, and may use the logistics management model to formulate a transportation plan for food. For example, the logistics management node 270 inputs the production quantity and production time of food obtained from the production management node 140 into the logistics management model. The logistics management node 270 then obtains the transportation quantity and transportation time of food from the logistics management model as a transportation plan. The logistics management node 270 then transports food based on the transportation quantity and transportation time of food obtained from the logistics management model. For example, the logistics management node 270 may output the transportation quantity and transportation time of food to a carrier.

[Hardware configuration]
Next, the hardware configuration of the node will be described. A node may be configured using a single hardware circuit or a combination of a hardware circuit and software. Alternatively, each node may be configured using multiple devices connected via a network. For example, a node may be configured using cloud computing. Alternatively, a node may be realized as a computer device including a computer (Central Processing Unit (CPU)), Read Only Memory (ROM), and Random Access Memory (RAM). In addition to the above configuration, a node may also be realized as a computer device including a Network Interface Circuit (NIC).

Figure 11 is a block diagram showing the configuration of an information processing device 600, which is an example of the hardware configuration of a node. The information processing device 600 includes a CPU 610, ROM 620, RAM 630, storage device 640, and NIC 650, and constitutes a computer device. The CPU 610 loads a program from at least one of the ROM 620 and storage device 640. The CPU 610 then controls the RAM 630, storage device 640, and NIC 650 based on the loaded program. The computer device including the CPU 610 then controls these components and realizes the functions of the behavior prediction node 110, food loss determination node 120, product management node 130, or production management node 140 in Figure 2.

When realizing each function, CPU 610 may use RAM 630 or storage device 640 as a temporary storage medium for programs and data. Furthermore, CPU 610 may use a recording medium reading device (not shown) to read a program contained in recording medium 690, which stores the program in a computer-readable format. Alternatively, CPU 610 may receive a program from an external device (not shown) via NIC 650, store the program in RAM 630 or storage device 640, and operate based on the stored program.

ROM 620 stores programs and fixed data executed by CPU 610. ROM 620 is, for example, a programmable ROM (P-ROM) or flash ROM. RAM 630 temporarily stores programs and data executed by CPU 610. RAM 630 is, for example, a dynamic RAM (D-RAM). Storage device 640 stores data and programs that the information processing device 600 stores long-term. Storage device 640 may also operate as a temporary storage device for CPU 610. Storage device 640 is, for example, a hard disk device, a magneto-optical disk device, a solid state drive (SSD), or a disk array device. ROM 620 and storage device 640 are non-volatile (non-transitory) storage media. On the other hand, RAM 630 is a volatile (transitory) storage media. The CPU 610 can operate based on a program stored in the ROM 620, the storage device 640, or the RAM 630. In other words, the CPU 610 can operate using a non-volatile recording medium or a volatile recording medium.

NIC 650 relays data exchange with other devices or nodes via a network (not shown). NIC 650 is, for example, a local area network (LAN) card. Furthermore, NIC 650 is not limited to being wired, and may also be wireless.

The information processing device 600 configured in this manner can realize functions similar to those of each node in Figure 2 based on a program. Note that the information processing device 600 may also realize the functions of the behavior prediction device 111 or loss determination device 121 in Figure 1. Alternatively, the information processing device 600 may realize the functions of each node included in Figure 10. Alternatively, the information processing device 600 may realize the functions of multiple devices or multiple nodes.

The present invention has been described above with reference to the embodiments, but the present invention is not limited to the above embodiments. Various modifications that would be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

10 Information processing system 11 Information processing system 12 Information processing system 110 Behavior prediction node 111 Behavior prediction device 120 Food loss determination node 121 Loss determination device 130 Product management node 140 Production management node 210 Personal information node 220 Behavior monitoring node 230 Biometric information measurement node 240 Product data node 251 Agriculture management node 252 Fisheries management node 253 Food production management node 254 Crop observation node 255 Fish catch quantity observation node 256 Food production monitoring node 260 Weather information node 600 Information processing device 610 CPU
620 ROM
630 RAM
640 Storage device 650 NIC
690 Recording Media

Claims (5)

a behavior prediction means for calculating a behavior probability that an individual will perform a predetermined behavior based on the individual's behavior history;
a loss determination means for calculating a loss probability of a predetermined item resulting from the action of the individual based on the action probability, determining whether or not the action probability needs to be changed based on the action probability and the loss probability, and outputting an instruction to change the action probability when it is determined that the action probability needs to be changed;
the behavior probability is a food purchase probability,
The loss probability is a food loss rate,
the individual is a consumer of a food product;
The behavior prediction means
The food purchase history includes at least past food information, personal information of the consumer, behavioral data of the consumer, and the food purchase probability,
storing a food purchase probability calculation model that is trained using at least the food information, the consumer's personal information, the consumer's behavioral data, and the food purchase probability included in the food purchase history, and that outputs the food purchase probability from at least the food information, the consumer's personal information, and the consumer's behavioral data;
Acquire food information and consumer personal information and behavioral data;
inputting the acquired food information, the personal information, and the behavioral data into the food purchase probability calculation model to calculate the food purchase probability;
outputting the acquired food information and personal information, the calculated food purchase probability, and the food purchase history to the loss determination means;
changing the food purchase probability in the food purchase history based on an instruction to change the food purchase probability obtained from the loss determination means;
outputting information related to food purchases to the consumer based on the instruction to change the food purchase probability;
The loss determination means
The food loss history includes at least the past food information, the personal information, and the food loss rate,
storing a food loss rate calculation model that calculates the food loss rate from at least the food information and the personal information, the food loss rate calculation model being trained using at least the food information, the personal information, and the food loss rate included in the food loss history;
acquiring the food information, the personal information of the consumer, the food purchase probability, and the food purchase history from the behavior prediction means;
The food loss rate calculation model calculates the food loss rate by inputting the acquired food information and the personal information,
Determine whether or not the food purchase probability needs to be changed based on the calculated food waste rate;
When it is determined that the food purchase probability needs to be changed, an instruction to change the food purchase probability is output to the behavior prediction means.
Information processing system. including product control measures for controlling food products sold to consumers;
the behavior prediction means outputs a price change instruction to the product management means based on the food purchase probability change instruction;
The product management means changes the price of the food based on the price change instruction acquired from the behavior prediction means.
The information processing system according to claim 1 . The loss determination means outputs an instruction to change the distribution quantity of food to the product management means based on the food loss rate;
The product management means changes the transportation quantity of the food based on the instruction to change the distribution quantity of the food obtained from the loss determination means.
The information processing system according to claim 2 . including production control measures to control the production of food;
The loss determination means outputs an instruction to change food production to the production management means based on the food loss rate,
The production management means changes the production quantity of food based on the instruction to change the production of food obtained from the loss determination means.
4. The information processing system according to claim 2 or 3 . The behavior prediction device
Calculating a behavior probability that the individual will perform a predetermined behavior based on the individual's behavior history;
The loss determination device
calculating a loss probability of a predetermined item associated with the action of the individual based on the action probability, determining whether or not the action probability needs to be changed based on the action probability and the loss probability, and outputting an instruction to change the action probability when it is determined that the action probability needs to be changed;
the behavior probability is a food purchase probability,
The loss probability is a food loss rate,
the individual is a consumer of a food product;
The behavior prediction device
The food purchase history includes at least past food information, personal information of the consumer, behavioral data of the consumer, and the food purchase probability,
storing a food purchase probability calculation model that has been trained using at least the food information, the consumer's personal information, the consumer's behavioral data, and the food purchase probability included in the food purchase history, and that outputs the food purchase probability from at least the food information, the consumer's personal information, and the consumer's behavioral data;
Acquire food information and consumer personal information and behavioral data;
inputting the acquired food information, the personal information, and the behavioral data into the food purchase probability calculation model to calculate the food purchase probability;
outputting the acquired food information and personal information, the calculated food purchase probability, and the food purchase history to the loss determination device;
changing the food purchase probability in the food purchase history based on an instruction to change the food purchase probability obtained from the loss determination device;
outputting information related to food purchases to the consumer based on the instruction to change the food purchase probability;
The loss determination device
The food loss history includes at least the past food information, the personal information, and the food loss rate,
storing a food loss rate calculation model that calculates the food loss rate from at least the food information and the personal information, the food loss rate calculation model being trained using at least the food information, the personal information, and the food loss rate included in the food loss history;
acquiring the food information, the personal information of the consumer, the food purchase probability, and the food purchase history from the behavior prediction device;
The food loss rate calculation model calculates the food loss rate by inputting the acquired food information and the personal information,
Determine whether or not the food purchase probability needs to be changed based on the calculated food waste rate;
When it is determined that the food purchase probability needs to be changed, an instruction to change the food purchase probability is output to the behavior prediction device.
Information processing methods.