WO2017145710A1

WO2017145710A1 - Information processing device, information processing method, and program

Info

Publication number: WO2017145710A1
Application number: PCT/JP2017/004041
Authority: WO
Inventors: 由幸小林; 正範勝; 文一中村
Original assignee: ソニー株式会社
Priority date: 2016-02-23
Filing date: 2017-02-03
Publication date: 2017-08-31

Abstract

[Problem] To improve livestock products by analyzing a causal factor and optimally adjusting the causal factor in order to obtain desired products related to stockbreeding. [Solution] An information processing device related to the present disclosure is provided with: an acquiring unit which acquires a plurality of pieces of stockbreeding-related information that pertains to stockbreeding; an accumulating unit which accumulates the stockbreeding-related information; and a cause-and-effect analyzing unit which analyzes, on the basis of the accumulated stockbreeding-related information, the relationship between an attention variable and a cause variable which is a cause of the attention variable.

Description

Information processing apparatus, information processing method, and program

The present disclosure relates to an information processing apparatus, an information processing method, and a program.

Traditionally, in the field of livestock farmers, for example, when raising cattle, by giving the optimum feed based on the farmer's experience, giving appropriate exercise, and optimally setting up environmental facilities such as cowshed, The desired cow is raised and shipped to a wholesaler.

In livestock breeding, there is a need to know the cause and effect, such as what can be seen to understand the physical condition of the livestock and what is a good quality meat quality indicator. Such indicators often differ depending on factors such as the characteristics of livestock individuals, farms, breeders, etc., making it difficult to obtain general knowledge. In addition, in order to find such knowledge, it was necessary for farmers to closely observe and consider livestock and their conditions. Even in that case, it was difficult to find such knowledge comprehensively, and there were cases where it was not possible to discover useful causes. It has been accumulated as the know-how of each livestock farmer and was sometimes lost without being inherited. This knowledge belonged to the farmer and was lost without being inherited.

For this reason, it has been desired to improve livestock products by analyzing factors that are factors for obtaining a desired animal product and optimally adjusting the factors.

According to the present disclosure, an acquisition unit that acquires a plurality of livestock related information related to livestock, a storage unit that stores the livestock related information, and a variable of interest and a factor of the variable of interest based on the stored livestock related information There is provided an information processing apparatus including a causal analysis unit that analyzes a relationship with a cause variable.

Further, according to the present disclosure, acquiring a plurality of livestock related information relating to livestock, storing the livestock related information, and based on the accumulated livestock related information, a variable of interest and a factor of the variable of interest An information processing method comprising: analyzing a relationship with a cause variable.

Further, according to the present disclosure, a means for acquiring a plurality of livestock related information relating to livestock, a means for storing the livestock related information, a cause that causes the variable of interest based on the stored livestock related information A program for causing a computer to function as a means for analyzing a relationship with a variable is provided.

As described above, according to the present disclosure, it is possible to improve livestock products by analyzing factors that are factors for obtaining a desired animal product and optimally adjusting the factors that are factors.
Note that the above effects are not necessarily limited, and any of the effects shown in the present specification, or other effects that can be grasped from the present specification, together with or in place of the above effects. May be played.

2 is a schematic diagram illustrating an information processing apparatus according to an embodiment of the present disclosure and a configuration around the information processing apparatus. FIG. It is a schematic diagram which shows the sensor used for the kind of sensing part, the sensing information detected by various sensors, the detail of sensing information, and sensing. It is a schematic diagram which shows the example of the information which an input information acquisition part acquires. It is a schematic diagram which shows the data stored in the database. It is a schematic diagram which shows the analysis result by a causal analysis part. FIG. 6 is a schematic diagram showing a directed graph when the user further specifies a variable “meat quality” from the state of FIG. 5. FIG. 10 is a characteristic diagram showing a result of analyzing a relationship between a noticed variable “wholesale price” and a variable “temperature” that is causal of “wholesale price”. It is a schematic diagram which shows the result of the causal analysis part performing the prediction display of the influence with respect to various variables. It is a schematic diagram which shows the directed graph at the time of designating an attention variable to "weight". FIG. 5 is a schematic diagram showing screens of “predicted performance display”, “cause variable list display”, and “cause variable influence display”. It is a flowchart which shows the process of the causal analysis by a causal analysis part. It is a schematic diagram which shows the mode of the discretization by a causal analysis part. It is a schematic diagram which shows UI for inputting cost. FIG. 5 is a characteristic diagram showing a relationship between a variable “cost” of interest and a variable “employee stress” that causes “cost”. It is a schematic diagram which shows the trade-off graph display of an effect and cost. It is a schematic diagram which shows the example which displayed the card | curd C shown in FIG. 15 as a list. It is a schematic diagram which shows the example which provides the function which can compare and examine how the attention variable changes in that case after the combination pattern of the variable which the user changed is input. It is a flowchart which shows the process for obtaining the causal analysis result of FIG. FIG. 10 is a schematic diagram illustrating an example in which an experiment proposal is made for a variable X whose dependency relationship does not appear in the directed graph even though the correlation with the variable of interest is high. It is a flowchart which shows the process of FIG. It is a flowchart which shows the process which estimates the magnitude | size of the influence of a variable. 16 is a flowchart showing processing performed by the Pareto optimal solution calculation unit 410 of the causal analysis unit 400 to perform the presentation of FIG. 15.

Hereinafter, preferred embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. In addition, in this specification and drawing, about the component which has the substantially same function structure, duplication description is abbreviate | omitted by attaching | subjecting the same code | symbol.

The description will be made in the following order.
1. 1. Configuration example of information processing apparatus 2. Outline of causal analysis Details of causal analysis

1. Example of configuration of information processing apparatus In the field of raising livestock, there is a need to know the cause and effect, such as what can be seen to understand the physical condition of the livestock and what is an index of good quality meat. In this embodiment, data of various sensors attached to livestock and various sensors installed in the farmer's environment are accumulated, causal analysis is performed on the data, and the analysis result is presented to the farmer who is a user.

First, with reference to FIG. 1, a configuration of an information processing apparatus 1000 according to an embodiment of the present disclosure and its surroundings will be described. As illustrated in FIG. 1, the information processing apparatus 1000 includes a sensing information acquisition unit 100, a feature extraction unit 200, a database 300, a causal analysis unit 400, an input information acquisition unit 500, an operation information acquisition unit 600, and a presentation information output unit 700. It is configured. Each component shown in FIG. 1 can be configured by a central processing unit such as a CPU and a program (software) for causing it to function.

The sensing information acquisition unit 100 acquires a sensor signal when a sensor (sensing unit 2000) attached to an individual animal, an environment where the animal is placed, a farmer, or the like senses time-series data, and the sensor signal Is stored in the database 300. The sensing unit 2000 includes various sensors.

FIG. 2 is a schematic diagram showing the type of the sensing unit 2000, sensing information detected by various sensors, details of the sensing information, and sensors used for sensing. As shown in FIG. 2, as a type of the sensing unit 2000, an individual wearing sensor to be attached to a livestock individual, an environmental sensor for obtaining environmental information such as a facility for raising livestock, a farm animal breeder, a robot, etc. Includes farmer-mounted sensors. As shown in FIG. 2, data to be sensed includes data sensed from livestock, data sensed from an environment such as a barn, and data sensed from a person such as a grower. Although the sensing unit 2000 obtains the sensing information, direct information acquisition, acquisition by wired communication, acquisition by wireless communication, and the like can be mentioned, but the method is not limited.

The input information acquisition unit 500 acquires information other than that obtained by sensing according to a user's operation input, and stores it in the database 300. The input information acquisition unit 500 acquires operation input information input by an input unit such as a keyboard or a mouse, for example. Examples of the method by which the input information acquisition unit 500 acquires operation input information include direct information acquisition, wired communication acquisition, and wireless communication acquisition, but the method is not limited.

FIG. 3 is a schematic diagram illustrating an example of information acquired by the input information acquisition unit 500. As shown in FIG. 3, the input information acquisition unit 500 acquires information related to livestock individuals and information related to farmers as shown in FIG. 3 and stores them in the database 300.

If there is information that cannot be used for causal analysis as it is among the information related to the sensor signal acquired by the sensing information acquisition unit 100 and the input information acquisition unit 500, the feature extraction unit 200 extracts the feature from the information and performs the causal analysis. The information is processed into usable information and stored in the database 300. For example, the feature extraction unit 200 calculates an amount of exercise (consumed calories) based on an individual's movement history and action type, and calculates an average value and a magnitude of change of each sensing information for each month.

The causal analysis unit 400 analyzes causality between variables for the data stored in the database 300. The causal analysis unit 400 performs causal analysis periodically or at a timing designated by the user. The analysis result by the causal analysis unit 400 is displayed as a directed graph or the like.

The causal analysis unit 400 includes a factor extraction unit 402, a directed graph estimation unit 404, a prediction performance analysis unit 406, an influence degree estimation unit 408, and a Pareto optimal solution calculation unit 410. Processing of each component of the causal analysis unit 400 will be described later.

The operation information acquisition unit 600 acquires information for operating the analysis result display by the causal analysis unit 400. For example, the operation information acquisition unit 600 acquires information input from an input unit such as a keyboard or a mouse. The method by which the operation information acquisition unit 600 acquires information includes direct information acquisition, acquisition by wired communication, acquisition by wireless communication, and the like, but the method is not limited. The presentation information output unit 700 outputs information for presenting the analysis result to the display device 3000.

2. Outline of Causal Analysis FIG. 4 is a schematic diagram showing data stored in the database 300. Here, a cow is illustrated as a livestock, and an example in which various variables associated with a cow individual ID are stored is shown. Variables include variables related to cow breeding and variables related to breeding results. As variables related to breeding, delivery workers, castration personnel, basic feed, feed composition, exercise amount, sleep time, number of chewing, breeding density, average air volume, average temperature, average humidity, average noise, average brightness, average number of pests, Examples include average breeding time, gender, frequent robot contact, and food consumption. Examples of the variables related to the growth result include body weight, carcass weight, carcass yield, loin core, back fat thickness, fat cross, meat color, meat fat color, meat quality, price, and the like.

The causal analysis unit 400 analyzes the causality between variables for the data shown in FIG. Specifically, the causal analysis unit 400 performs a causal analysis using, for example, a known max-min Hill-Climbing method, and analyzes the causality between variables. FIG. 5 is a schematic diagram illustrating an example of the analysis result. The presentation information output unit 700 outputs information for presenting the analysis result by the causal analysis unit 400 to the display device 3000, whereby the analysis result illustrated in FIG. 5 is displayed on the display device 3000.

In FIG. 5, a portion where each variable is connected by a line indicates that there is a cause and effect between the variables. As described above, when the user inputs with an input unit such as a keyboard or a mouse, the operation information acquisition unit 600 acquires information for operating the analysis result display by the causal analysis unit 400. FIG. 6 is a schematic diagram (directed graph) showing a case where the user further specifies the variable “meat quality” from the state of FIG. 5. The useful graph estimation unit 404 of the causal analysis unit 400 analyzes the attention variable “meat” and the attention variable “meat” as shown in FIG. 6 based on the analysis result of the designated variable “meat quality” and the causal variable. A useful graph in which cause variables (the amount of contact with the robot, the amount of exercise, the noise, the amount of meal) associated with the above are linked is estimated. The presentation information output unit 700 outputs information for presenting the analysis result by the causal analysis unit 400 to the display device 3000, whereby the analysis result illustrated in FIG. 6 is displayed on the display device 3000.

According to the analysis results shown in FIG. 6, the “meat quality” of cattle is due to the large number of contact with the robot, the amount of exercise, the noise, and the amount of food. Therefore, changing these variables changes the meat quality. I understand that. Therefore, in order to improve the “meat quality”, the user can take measures such as changing variables such as exercise amount and meal amount. Therefore, the user can understand the cause that has not been noticed so far in order to improve the meat quality, and can take a new countermeasure as a result.

Furthermore, according to the analysis by the influence degree estimation unit 408 of the causal analysis unit 400, it is possible to present the magnitude of the influence on the variable of interest among the causal variables. In the example of FIG. 6, in order of the magnitude of the influence on “meat quality”, the first place is presented as the amount of exercise (30%), the second place is the amount of meal (15%), and the third place is the amount of noise (6%). You can also. Thereby, the user can recognize that changing the amount of exercise is the most effective for improving the “meat quality”. FIG. 21 is a flowchart illustrating processing in which the influence degree estimation unit 408 estimates the magnitude of the influence. First, in step S40, the user designates an attention variable. In the next step S42, the variables that cause the variable of interest are listed. In the next step S44, the variables of interest are sorted in order of influence. In the next step S46, a causal variable is presented.

Similarly, when paying attention to other variables, by designating that variable, the specified variable and the causal variable are analyzed and displayed as in FIG. Thus, for example, when a user inputs a variable (health condition, meat quality, etc.) that the user wants to improve, the factor (causal) that is an analysis result is presented. Therefore, the user can change the noted variable by changing a variable having a factor (causal). Thereby, based on the analysis result by the causal analysis unit 400, the user can optimally change various variables when raising the cow. Since it can be understood which factors have an influence on indicators such as physical condition and meat quality of livestock, if the cause is known, measures can be taken to improve the indicators.

FIG. 7 is a characteristic diagram showing the result of analyzing the relationship between the noticed variable “Wholesale Price” and the variable “Temperature” causal of “Wholesale Price”. In this way, the variable of interest and the relationship between the variable of interest and the causal variable can also be shown in a graph. The presentation information output unit 700 outputs information for presenting the analysis result of FIG. 7 to the display device 3000, whereby the analysis result shown in FIG. 7 is displayed on the display device 3000. Thereby, the user can recognize that there is a possibility that the “wholesale price” can be increased by setting the temperature to around 20 degrees.

The information shown in FIGS. 6 and 7 may be presented in text. For example, if you increase the number of contact with cattle by 10%, the chance of getting sick will be reduced by 5%, or you may want to reduce the noise. A sentence from the viewpoint of (cow) may be presented. Documenting may be performed by a technique such as a recurrent neural network (RNN). Further, these sentences may be synthesized by speech and uttered.

The presentation of causal information recognizes user behavior and presents it at an appropriate time. For example, at the timing when the user feeds the cow, the wholesale price may increase if the amount is smaller than usual. Alternatively, the information may be presented in a fortune-telling style, such as “Today's lucky food is hay, lucky temperature is 18 degrees”.

FIG. 8 is a schematic diagram showing a result of the causal analysis unit 400 performing prediction display of influences on various variables. As shown in FIG. 8, when an action is taken, predictions about how the action affects a plurality of variables can be presented in a list. In the example of FIG. 8, when the food is changed from the current food to food A, the meat quality changes by 5%, the disease probability changes by 3%, and the cost changes by 2%.

Also, if the causal analysis result is ambiguous, the farmer may be encouraged to take a sensing action to clarify the result. For example, information such as “There is a possibility that the influence of the color may be understood if the color of the barn is changed” is presented to the user. It is possible to provide a service that reflects user behavior in sensing, accumulates the sensed data, and performs further analysis and presents the data.

Furthermore, the information processing device 1000 is connected to a device that actually controls the variables (for example, an air conditioner that controls the room temperature in the barn or a device that mixes food), and the variable that causes the factor is based on the causal analysis result. It is also possible to perform automatic control so that changes directly.

3. About the details of causal analysis When the operation information acquisition unit 600 acquires information for operating the display of the analysis result, the display of the analysis result transitions, and a more detailed analysis result can be obtained. Here, it is assumed that the variable of interest is designated as “weight”. As a result, “basic feed ID”, “feed formulation ID”, “sleeping time”, and “breeding density” are extracted as variables that cause “weight” by the causal analysis unit 400, and the directed graph shown in FIG. 9 is obtained. It is done. When the user performs a predetermined operation, the screen transitions from the state shown in FIG. 9 to each screen of “predicted performance display”, “cause variable list display”, and “cause variable influence display” shown in FIG. be able to. The analysis result of each screen can be obtained by performing the influence analysis of the factor on the result by generalized linear regression.

In the “predicted performance display” shown in FIG. 10, the body weight predicted on the basis of four variables (basic feed ID, feed composition ID, sleep time, breeding density) and “cause weight” (on the horizontal axis of the predicted performance display). "Weight" shown) and the actual weight are shown, and in this example, the correlation degree R is 0.89. Therefore, it can be seen that the weight can be predicted with high accuracy. When the predicted body weight 'matches the actual body weight, the correlation degree R is 1.0. This prediction is performed by the prediction unit 402 of the causal analysis unit 400.

As shown in FIG. 10, in the “list display of causal variables”, the degree of influence of causal variables is displayed. As for the degree of influence, “basic feed ID” is the highest at 78.4%, “feed formulation ID” is 20.3%, “sleeping time” is 0.9%, and “breeding density” is 0.4%. . Therefore, “basic feed ID” and “feed formulation ID” have a large influence on “body weight”, and it can be seen that there is a possibility that weight can be improved by controlling these variables. In particular, since the degree of influence of “basic feed ID” is the largest, it is most effective to change “basic feed ID” in order to change “weight”.

On the “Indication of the influence of the causal variable” screen, for each of the two variables (“basic feed ID” and “feed formulation ID”) that greatly affect “weight”, which range of the variable is “weight” It shows whether the impact on In the case of the basic feed ID, it can be seen that the weight can be increased by using the basic feed with

ID

2 or 3. Therefore, it can be understood that the user only needs to give a basic feed having an ID of 2 to 3 in order to increase the “weight”. This analysis is performed by the prediction performance analysis unit 406.

On the other hand, in the case of feed blending ID, it can be seen that weight is increased by blending with ID of 5.0 or more. Therefore, it can be understood that the user only needs to blend with a blend having an ID of 5.0 or more in order to increase the “weight”.

FIG. 11 is a flowchart showing a causal analysis process performed by the causal analysis unit 400. First, in step S <b> 10, a causal analysis target data set is input to the causal analysis unit 400. As a result, the data set shown in FIG. 4 is input to the causal analysis unit 400.

In the next step S12, the continuous value variable is discretized for the input data. Only the discretized variables (variables having specific values such as 0, 1, 2, 3) that can be handled by the Max-min Hill climbing method described above are continuous values (such as 0.0 to 1.0). The variable with a continuous value of () cannot be handled, so discretization is performed. In the present embodiment, the interval between the minimum value and the maximum value is evenly discretized to a predetermined number. FIG. 12 is a schematic diagram showing the state of the discretization process. When the continuous value is discretized in 8 steps (0, 1, 2,..., 7), the minimum value and the maximum value are divided into eight.

In the next step S14, the DAG (directed graph) is estimated by the Max-min Hill climbing method. Thereby, for each variable of the data set, an estimation result as to which other variable is a factor can be obtained, and a directed graph as shown in FIG. 6 can be obtained.

In the next step S16, the influence analysis of each factor is performed by generalized linear regression. Thereby, the relationship between the variable and the variable that causes it is obtained, and “prediction performance display”, “list display of the cause variable”, and “effect display of the cause variable” as shown in FIG. Can get a relationship.

FIG. 13 and FIG. 14 are schematic diagrams showing an example in which the system can automatically estimate the cost fluctuation due to the fluctuation of the variable for the variable that can be controlled among the variables and the cost that is difficult to calculate directly. . Here, FIG. 13 illustrates a UI for inputting a cost, and information input via the UI is acquired by the operation information acquisition unit 600. FIG. 14 is a characteristic diagram showing the relationship between the variable “cost” of interest and the variable “employee stress” that causes “cost”. Here, a causal analysis result that “cost” decreases as “employee stress” increases is shown. By using the result of causal analysis, it is possible to analyze the relationship between the variable “cost” and other variables, and to estimate the effect on “cost” when a certain variable is changed. At this time, estimation in consideration of the value (trivial cost) input through the cost input UI is possible.

FIG. 15 is a schematic diagram showing a trade-off graph display of effect and cost. As shown in FIG. 15, for example, the measure plan is presented in a scatter diagram with the vertical axis representing the effect (cow price) and the horizontal axis representing the running cost. The measure plan to be presented is selected and presented as Pareto optimal (better than other solutions in either effect or cost) in two variables of effect and cost.

In the example shown in FIG. 15, the running cost and the price change when the variable “food” that causes the cost is changed. It can be seen that changing the food to food A for the current situation (NOW) increases the running cost but also the price of the cow. It can also be seen that changing the food to the food B for the current situation reduces the running cost but also the price of the cow. The user can recognize the merits and demerits of changing the food based on the analysis result of FIG. In FIG. 15, the measure plan that is not Pareto optimal is not displayed or the display priority is lowered.

FIG. 16 is a schematic diagram showing an example in which the cards C shown in FIG. 15 are displayed as a list. One card describes one measure and the effects and costs obtained thereby. Cards can be sorted and displayed based on cost and effectiveness.

FIG. 22 is a flowchart showing a process performed by the Pareto optimal solution calculation unit 410 of the causal analysis unit 400 to perform the presentation of FIG. First, in step S50, the cause variable-attention variable-cost relationship is listed while changing the cause variable.

In the next step S52, the Pareto optimal solutions of the attention variable-cost relationship are listed. In the next step S54, a measure plan is presented by the trade-off graph shown in FIG. 15 or the display by the card C shown in FIG. In the next step S56, the user selects and adopts one of the measure plans.

FIG. 17 is a schematic diagram showing an example of providing a function capable of comparing and examining how the variable of interest changes in such a case when a combination pattern of variables changed by the user is input. In FIG. 17, variables shown in bold indicate numerical values input and changed by the user. The underlined variable indicates a variable whose value has changed due to the changed variable. The influence estimation unit 408 of the causal analysis unit 400 estimates changes in other variables based on the numerical values changed by the user shown in bold. The user can recognize a variable whose value has been changed by changing the desired variable as appropriate. It is also possible to create and save a plurality of plans as shown in FIG.

FIG. 18 is a flowchart showing a process for obtaining the causal analysis result of FIG. First, in step S20, the user inputs a cause variable after intervention. As a result, the variables shown in bold in FIG. 17 are changed. In the next step S22, the causal analysis unit estimates the values of other variables that are affected by the changed cause variable. Thereby, the value of the variable underlined in FIG. 17 is estimated. In the next step S24, a list of variable values is displayed. In the next step S26, the changed content is saved as a plan. Thus, for example, plan C is stored in the list shown in FIG. In the next step S28, a plurality of plans stored in the past are presented. As a result, the plans A and B saved in the past are presented together with the plan C as shown in FIG.

As described above, when there is a variable with a small amount of data and a causal analysis result that is ambiguous, it is possible to encourage the user to take an action to acquire the variable. FIG. 19 is a schematic diagram showing an example in which an experiment proposal is made for a variable X whose dependency relationship does not appear in the directed graph even though the correlation with the target variable is high.

As shown in FIG. 19, it is assumed that the variable B is set as the variable of interest Y, and a high correlation with the variables A, C, and D can be confirmed. Here, when the variable A is a variable X for which an experiment proposal is made, a path of X → Y exists in the directed graph. Accordingly, no further change proposal (experimental proposal) is made for the variable A in which a path to the target variable Y exists.

Similarly, when the variable C is a variable X for which an experiment is proposed, there is a Y → X path in the directed graph. Accordingly, no further change proposal (experimental proposal) is made for the variable C in which a path to the target variable Y exists.

On the other hand, if the variable D is the variable X for which an experiment is proposed, there is no Y → X or X → Y path in the directed graph. Therefore, for the variable D for which there is no path to the target variable Y, a further change proposal (experimental proposal) is made.

FIG. 20 is a flowchart showing the processing of FIG. First, in step S30, other variables X having a high correlation with the target variable Y are listed. In the next step S32, it is determined whether or not a path of X → Y or Y → X exists in the directed graph. If there is no path, the process proceeds to step S34. In step S34, the acquisition of data with the variable X condition changed is proposed. On the other hand, if a path of X → Y or Y → X exists in the directed graph, the process ends without proposing acquisition of data.

If the user actually observes as suggested, the variable will automatically be removed from the proposal. When the correlation decreases as a result of increasing the data, it becomes clear that there is no cause and effect between the variable D and the variable B. In this case, it becomes out of the proposal when enumerating highly correlated variables. If the correlation remains even if the data is increased, some causality exists between the variable D and the variable B. Therefore, the path of the variable D → the variable B or the variable B → the variable D is present in the directed graph at the time of the causal analysis. Will appear. In this case, it is excluded from the proposal when determining whether or not an X → Y or Y → X path exists.

As described above, according to the present embodiment, a plurality of data relating to livestock is accumulated in the database 300, and based on the accumulated data, the relationship between the attention variable and the cause variable that causes the attention variable is analyzed. To be presented to the user. Thereby, the user can understand the cause for improving the attention variable, and can take a new countermeasure as a result.

The preferred embodiments of the present disclosure have been described in detail above with reference to the accompanying drawings, but the technical scope of the present disclosure is not limited to such examples. It is obvious that a person having ordinary knowledge in the technical field of the present disclosure can come up with various changes or modifications within the scope of the technical idea described in the claims. Of course, it is understood that it belongs to the technical scope of the present disclosure.

In addition, the effects described in this specification are merely illustrative or illustrative, and are not limited. That is, the technology according to the present disclosure can exhibit other effects that are apparent to those skilled in the art from the description of the present specification in addition to or instead of the above effects.

The following configurations also belong to the technical scope of the present disclosure.
(1) An acquisition unit that acquires a plurality of livestock related information related to livestock,
An accumulating unit for accumulating the livestock related information;
Based on the accumulated livestock related information, a causal analysis unit that analyzes the relationship between the attention variable and the cause variable that causes the attention variable;
An information processing apparatus comprising:
(2) The information processing apparatus according to (1), wherein the causal analysis unit includes a directed graph estimation unit that estimates a directed graph indicating a relationship between the attention variable and the cause variable.
(3) The information processing apparatus according to (1) or (2), wherein the causal analysis unit includes an influence degree estimation unit that estimates an influence degree of the cause variable with respect to the attention variable.
(4) The information processing apparatus according to (3), wherein the influence degree estimation unit estimates a change in the attention variable or another cause variable when an arbitrary cause variable is changed.
(5) The information processing apparatus according to any one of (1) to (4), wherein the causal analysis unit includes a prediction unit that predicts the attention variable from the cause variable.
(6) The causal analysis unit includes a Pareto optimal solution calculation unit that calculates a Pareto optimal solution for a plurality of patterns in which the variable of interest changes when the cause variable is changed. ).
(7) The information processing apparatus according to any one of (1) to (6), wherein the livestock related information is sensing information detected by a sensor attached to a livestock subject to livestock.
(8) The information processing apparatus according to any one of (1) to (7), wherein the livestock related information is sensing information in which a sensor detects a living environment of livestock that is a subject of livestock.
(9) The information processing apparatus according to any one of (1) to (7), wherein the livestock related information is information input by a livestock farmer.
(10) The information processing apparatus according to any one of (1) to (9), further including an operation information acquisition unit that acquires operation information for designating an analysis result by the causal analysis unit.
(11) The information processing apparatus according to any one of (1) to (10), further including a presentation information output unit that outputs presentation information for presenting an analysis result by the causal analysis unit.
(12) The information processing apparatus according to any one of (1) to (11), further including a feature extraction unit that extracts a feature amount from the livestock-related information and stores the feature amount in the storage unit.
(13) The information processing apparatus according to any one of (1) to (12), further prompting acquisition of the livestock related information based on a degree of association between the attention variable and an arbitrary cause variable.
(14) obtaining a plurality of livestock related information relating to livestock;
Storing the livestock related information;
Analyzing the relationship between the variable of interest and the causal variable that causes the variable of interest based on the accumulated livestock related information;
An information processing method comprising:
(15) Means for obtaining a plurality of livestock related information relating to livestock,
Means for accumulating the livestock related information,
Means for analyzing the relationship between the attention variable and the causal variable that causes the attention variable based on the accumulated stock raising-related information;
As a program to make the computer function as.

DESCRIPTION OF SYMBOLS 1000 Information processing apparatus 100 Sensing information acquisition part 200 Feature extraction part 300 Database 400 Causal analysis part 402 Prediction part 404 Useful graph estimation part 406 Predictive performance analysis part 408 Influence degree estimation part 410 Pareto optimal solution calculation part 500 Input information acquisition part 600 Operation Information acquisition unit 700 Presentation information output unit

Claims

An acquisition unit for acquiring a plurality of livestock related information related to livestock;
An accumulating unit for accumulating the livestock related information;
Based on the accumulated livestock related information, a causal analysis unit that analyzes the relationship between the attention variable and the cause variable that causes the attention variable;
An information processing apparatus comprising:
The information processing apparatus according to claim 1, wherein the causal analysis unit includes a directed graph estimation unit that estimates a directed graph indicating a relationship between the attention variable and the cause variable.
The information processing apparatus according to claim 1, wherein the causal analysis unit includes an influence degree estimation unit that estimates an influence degree of the cause variable with respect to the attention variable.
The information processing apparatus according to claim 3, wherein the influence degree estimation unit estimates a change in the attention variable or another cause variable when an arbitrary cause variable is changed.
The information processing apparatus according to claim 1, wherein the causal analysis unit includes a prediction unit that predicts the attention variable from the cause variable.
The information processing apparatus according to claim 1, wherein the causal analysis unit includes a Pareto optimal solution calculation unit that calculates a Pareto optimal solution for a plurality of patterns in which the variable of interest changes when the cause variable is changed.
The information processing apparatus according to claim 1, wherein the livestock related information is sensing information detected by a sensor attached to livestock that is a target of livestock.
The information processing apparatus according to claim 1, wherein the livestock related information is sensing information in which a sensor detects a living environment of livestock that is a target of livestock.
The information processing apparatus according to claim 1, wherein the livestock related information is information input by a livestock farmer.
The information processing apparatus according to claim 1, further comprising an operation information acquisition unit that acquires operation information for designating an analysis result by the causal analysis unit.
The information processing apparatus according to claim 1, further comprising: a presentation information output unit that outputs presentation information for presenting an analysis result by the causal analysis unit.
The information processing apparatus according to claim 1, further comprising a feature extraction unit that extracts a feature amount from the livestock-related information and accumulates the feature amount in the accumulation unit.
The information processing apparatus according to claim 1, further prompting acquisition of the livestock related information based on a degree of association between the attention variable and an arbitrary cause variable.
Obtaining multiple livestock related information on livestock;
Storing the livestock related information;
Analyzing the relationship between the variable of interest and the causal variable that causes the variable of interest based on the accumulated livestock related information;
An information processing method comprising:
Means for obtaining a plurality of livestock related information relating to livestock;
Means for accumulating the livestock related information,
Means for analyzing the relationship between the attention variable and the causal variable that causes the attention variable based on the accumulated stock raising-related information;
As a program to make the computer function as.