Breeding support device, breeding support method, and breeding support program

Abstract

In order to assist breeding suitably, a selective breeding assistance device (1) comprises: a reception unit (11) for receiving a request concerning breeding; a generation unit (12) for, by using a trained model obtained by learning relationships each between a characteristic of an existing breed and the breeding process for the existing breed, generating, on the basis of the request, response information including information concerning an existing breed that serves as a crossbreeding candidate for developing a new breed; and an output unit (13) for outputting the response information.

Description

[Technical Field]

[0001] The present invention relates to a plant breeding support device and the like that generates information related to breeding. [Background technology]

[0002] Breeding to give plants and animals more desirable traits has been practiced for a long time, mainly in the fields of agriculture, forestry, and livestock. Breeding involves the process of selecting varieties to crossbreed, crossbreeding the selected varieties, and verifying whether the new varieties resulting from the crossbreeding possess the desired traits (see, for example, Patent Document 1 listed below). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2013-55963 Summary of the Invention [Problem to be solved by the invention]

[0004] Breeding requires the above process to be repeated until a new variety with the desired traits is found, which requires a great deal of time and effort, and there is a need for technologies to assist in breeding.

[0005] One aspect of the present invention has been made in view of the above-mentioned problems, and one of its objectives is to provide a technique for suitably supporting breeding. [Means for solving the problem]

[0006] A plant breeding support device according to one aspect of the present invention comprises a receiving means for receiving a request for breeding, a generating means for generating response information including information about existing varieties that are candidates for breeding to develop a new variety based on the request using a trained model that has learned the relationship between the properties of existing varieties and the breeding process of the existing varieties, and an output means for outputting the response information.

[0007] In one aspect of the present invention, a method for supporting plant breeding involves a computer receiving a request for breeding, and using a trained model that has learned the properties of existing varieties and the relationships between the breeding processes of those existing varieties, generating response information based on the request that includes information about existing varieties that are candidates for breeding to develop a new variety, and outputting the response information.

[0008] A variety improvement support program according to one aspect of the present invention causes a computer to perform the following processes: accepting a breeding request; generating response information based on the request using a trained model that has learned the relationship between the properties of existing varieties and the breeding process of the existing varieties, including information about existing varieties that are candidates for breeding to develop a new variety; and outputting the response information. [Effects of the Invention]

[0009] According to one aspect of the present invention, breeding can be favorably supported. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a block diagram showing a configuration of a plant breeding support device according to a first exemplary embodiment of the present invention. [Figure 2] FIG. 1 is a flowchart showing the flow of a plant breeding support method according to a first exemplary embodiment of the present invention. [Figure 3] FIG. 1 is a diagram illustrating feature learning in graph-based relational learning. [Figure 4] FIG. 10 is a diagram showing an overview of a plant breeding support method according to a second exemplary embodiment of the present invention. [Figure 5] FIG. 10 is a block diagram showing the configuration of a plant breeding support device according to a second exemplary embodiment of the present invention. [Figure 6] FIG. 10 is a flowchart showing the flow of processing executed by a plant breeding support device according to a second exemplary embodiment of the present invention. [Figure 7] FIG. 10 is a diagram illustrating an example of response information. [Figure 8] FIG. 10 is a diagram showing an overview of a plant breeding support method according to a third exemplary embodiment of the present invention. [Figure 9] FIG. 10 is a block diagram showing the configuration of a plant breeding support device according to a third exemplary embodiment of the present invention. [Figure 10] FIG. 10 is a flowchart showing the flow of processing executed by a plant breeding support device according to a third exemplary embodiment of the present invention. [Figure 11] FIG. 1 is a diagram showing an outline of a plant breeding support method including a process for identifying varieties similar to a base variety. [Figure 12] FIG. 10 is a flowchart showing the flow of processing executed by the plant breeding support device when identifying a variety similar to a base variety. [Figure 13] FIG. 10 is a diagram showing an overview of a plant breeding support method according to a fourth exemplary embodiment of the present invention. [Figure 14] FIG. 10 is a block diagram showing the configuration of a plant breeding support device according to a fourth exemplary embodiment of the present invention. [Figure 15] FIG. 10 is a flowchart showing a flow of processing executed by a plant breeding support device according to a fourth exemplary embodiment of the present invention. [Figure 16] FIG. 10 is a diagram showing an overview of a plant breeding support method according to a fifth exemplary embodiment of the present invention. [Figure 17] FIG. 10 is a block diagram showing the configuration of a plant breeding support device according to a fifth exemplary embodiment of the present invention. [Figure 18] FIG. 11 is a flowchart showing a flow of processing executed by a plant breeding support device according to a fifth exemplary embodiment of the present invention. [Figure 19] FIG. 10 is a diagram illustrating an example of predicting the traits of a new variety based on feature amounts calculated from a new variety graph and an existing variety graph. [Figure 20] FIG. 1 is a configuration diagram for realizing a breeding support device by software. DETAILED DESCRIPTION OF THE INVENTION

[0011] Exemplary Embodiment 1 A first exemplary embodiment of the present invention will be described in detail with reference to the drawings. This exemplary embodiment is a basic form of the exemplary embodiments described below.

[0012] (Variety improvement support device) The configuration of a plant breeding support device 1 according to this exemplary embodiment will be described with reference to Fig. 1. Fig. 1 is a block diagram showing the configuration of the plant breeding support device 1. As shown in the figure, the plant breeding support device 1 includes a receiving unit (receiving means) 11, a generating unit (generating means) 12, and an output unit (output means) 13.

[0013] The receiving unit 11 receives a request for breeding. The generating unit 12 generates response information based on the request, including information about existing varieties that are candidates for breeding to develop a new variety, using a trained model that has learned the properties of existing varieties and the relationship between the breeding processes of the existing varieties. The output unit 13 outputs the response information.

[0014] The "traits of existing varieties" refer to the properties of existing varieties, and include, for example, properties related to the growing environment, such as drought tolerance, high temperature tolerance, and low temperature tolerance, as well as properties of the harvested product, such as high yield (high harvest volume) and good eating quality, and disease resistance. The "breeding process of existing varieties" refers to how the existing variety was developed, and includes, for example, the parent varieties, the number of crosses made to create the existing variety, and traits acquired or lost during crossing. Traits refer to morphological or physiological properties.

[0015] The plant breeding support device 1 having the above configuration can generate useful response information including information on existing varieties that are candidates for breeding to develop new varieties. Therefore, the above configuration has the effect of providing effective support for breeding.

[0016] (Breeding Support Program) The functions of the above-described plant breeding support device 1 can also be realized by a program. The plant breeding support program according to this exemplary embodiment causes a computer to execute the following processes: accepting a request for breeding; generating response information including information about existing varieties that are candidates for breeding to develop a new variety based on the request using a trained model that has learned the relationship between the properties of existing varieties and the breeding process of the existing varieties; and outputting the response information. This plant breeding support program provides the effect of providing effective support for breeding.

[0017] (Variety improvement support method) The plant breeding support method according to this exemplary embodiment will be described with reference to Fig. 2. Fig. 2 is a flow chart showing the flow of the plant breeding support method according to the first exemplary embodiment of the present invention.

[0018] In S11, the computer accepts a breeding request. The request may be accepted via any input device. For example, the request may be accepted via a mouse, keyboard, touch panel, or voice input device.

[0019] In S12, the computer uses a trained model that has learned the properties of existing varieties and the relationship between the breeding processes of those existing varieties to generate response information including information about existing varieties that are candidates for breeding to develop a new variety, based on the request received in S11.

[0020] In S13, the computer outputs the response information generated in S12. The output device can be any device, and for example, the information may be output to a display device to display the information, or to an audio output device to output the information as audio.

[0021] As described above, in the plant breeding support method according to this exemplary embodiment, a computer receives a request for breeding (S11), generates response information including information about existing varieties that are candidates for breeding to develop a new variety based on the request received in S11 using a trained model that has learned the properties of existing varieties and the relationships between the breeding processes of those existing varieties (S12), and outputs the response information generated in S12 (S13). This plant breeding support method has the effect of providing appropriate support for breeding.

[0022] Note that each step in the above-described breeding support method may be executed by one computer (e.g., the breeding support device 1), or each step may be executed by a different computer. This also applies to the flows described in the second and subsequent exemplary embodiments.

[0023] [Graphs and Learning] In the following, a graph, which is an example of information that can be used to support breeding in the first exemplary embodiment and the following exemplary embodiments (hereinafter referred to as each exemplary embodiment), will be described. Learning of the graph and prediction using the graph will also be described.

[0024] (graph) A graph here refers to data with a structure consisting of multiple nodes and links connecting the nodes. The type of link that represents the relationship between nodes is also called a "relation." Links are also sometimes called edges. Graphs can be broadly divided into directed graphs, in which each link has a direction, and undirected graphs, in which each link does not have a direction. It is possible to use either directed graphs or undirected graphs, or to use a combination of the two.

[0025] In each exemplary embodiment, when a graph is used, the nodes may represent tangible or intangible elements related to breeding. For example, - Variety identification information (e.g. variety name or ID) - Traits of the variety (e.g. disease resistance, heat resistance, cold resistance, high yield, taste of the harvested product, etc.) Genome information Classification as breeding material (for example, intermediate parent plants with excellent genetic characteristics but shortcomings as practical varieties, introduced lines that are strains imported from foreign countries, native varieties that are strains developed and maintained in ancient times, etc.) It is possible to use a graph containing nodes representing various elements such as the above. Note that the graph may contain multiple nodes corresponding to one element. For example, since there are two breeding parents for a certain variety, the node representing the breeding parents is represented by two separate nodes. The same applies to other elements.

[0026] If there are nodes corresponding to the elements above, the links connecting such nodes are The relationship between one breed and another The relationship between an element and the numerical value of that element The relationship in which a certain element possesses a certain trait For example, a link connecting a node representing a certain variety with a node representing another variety may represent a relationship in which the certain variety is a parent of the other variety.

[0027] (Learning and Prediction) For graphs such as those described above, graph-based relationship learning can be performed by applying machine learning techniques. Such learning enables classification and prediction processes to be performed using graphs. Note that in each exemplary embodiment, such learning may be performed as part of breeding support, or a trained graph that has already undergone such learning may be used.

[0028] In graph-based relationship learning, first, the feature of each node is calculated. The feature may be in vector format, for example. By expressing the feature of each node as a feature vector, it becomes possible to learn about graphs in which nodes of various formats coexist. For example, graph-based relationship learning can also be performed on graphs that include images, numerical values, and the like that indicate the various elements described above.

[0029] Next, the feature values ​​of each node are updated based on the links connected to each node and the nodes to which those links are connected. This process is similar to the convolution process in a convolutional neural network. This will be explained with reference to Figure 3. Figure 3 is a diagram explaining feature learning in graph-based relational learning.

[0030] The graph shown in Fig. 3 includes four nodes A to D. Node A is connected to nodes B and C, and node C is connected to node D. After calculating the initial features of these four nodes, multiple convolutions are performed as described below to update the features of each node.

[0031] In the first convolution, the initial feature of node A is multiplied by a predetermined weight and added to the feature of nodes B and C connected to node A. For node C, the initial feature of node C is multiplied by a predetermined weight and added to the feature of node D. Directed In the case of a graph, weights are adjusted according to the direction of the links.

[0032] In the second convolution, as in the first convolution, the feature of each node is multiplied by a predetermined weight and then added with the feature of the node linked to that node. Here, the feature of node C reflects the feature of node D due to the first convolution. Therefore, in the second convolution, not only the feature of node C but also the feature of node D is reflected in node A.

[0033] By repeating the above process a number of times according to the node hierarchy, the features of each node directly or indirectly connected by links are mutually reflected. In graph-based relationship learning, the weight values ​​used for the above weighting are optimized based on the known relationships between nodes. By using such a trained graph (which can also be called a trained model), it is possible to predict the relationships between nodes and the nodes to which links will lead, as described below.

[0034] (Inter-node relationship prediction) By performing the above-described learning, it becomes possible to predict relationships between nodes that are not explicitly stated in the original graph. To perform node-to-node relationship prediction, a user simply specifies two nodes and requests that the relationship between those nodes be returned. For example, if a user inputs a request asking about the relationship between the "variety A" node and the "variety B" node, node-to-node relationship prediction can predict whether the relationship, i.e., the link, connecting these nodes is a "good mating partner." In addition, node-to-node relationship prediction can also calculate the probability (likelihood) of the predicted result. The same applies to node prediction, which will be described below.

[0035] (node ​​prediction) Furthermore, by performing the above-described learning, it is also possible to predict the nodes connected to a given node via a specified link. To perform node prediction, a user simply specifies a node and a link starting from that node, and requests that the linked node be returned. For example, suppose a user inputs a request asking about the nodes connected to the "variety A" node via a "trait" link. In this case, node prediction can predict whether the node connected to the "variety A" node via the "trait" link has "good eating quality" or "disease resistance," for example.

[0036] Exemplary Embodiment 2 (overview) 4 is a diagram showing an overview of a breeding support method according to this exemplary embodiment. In this exemplary embodiment, an example of supporting breeding using a base breeding graph and an existing breeding graph will be described.

[0037] The existing variety graph includes multiple nodes related to existing varieties and links indicating the relationships between the nodes, and is a learned graph of the relationships between the nodes, and is a learned model. The existing variety graph can also be called a knowledge graph. Note that a collection of nodes and links corresponding to one existing variety may be called an existing variety graph, or a collection of nodes and links corresponding to multiple existing varieties may be called an existing variety graph collectively.

[0038] For example, in Figure 4, the graph containing the nodes "Variety A" to "Variety C" is the existing variety graph. This existing variety graph includes nodes indicating that the parents (meaning the breeding parents; the same applies below) of the existing variety variety A are "Variety a1" and "Variety a2," nodes indicating that variety A has the traits of "high yield" and "good eating taste," and nodes indicating that breeding of variety A required "a3" crossbreedings. The existing variety graph in Figure 4 also indicates that the parents of "Variety C" are "Variety A" and "Variety B," and that "Variety A" and "Variety B" are "poor breeding partners." In this way, the existing variety graph shown in Figure 4 is the result of learning from examples of crossbreeding between existing varieties.

[0039] It is also possible to learn links between breeds that are "good mating partners" and not learn links between breeds that are "unsuitable mating partners" (as negative examples). However, learning links between breeds that are "unsuitable mating partners" has the advantage of being useful for predicting positive examples (good mating partners in this example) and for predicting risks.

[0040] The base variety graph is a graph that includes multiple nodes related to the base variety, which is one of the breeding parents of the new variety to be bred. In Figure 4, the graph that includes the node "base variety" is the base variety graph. This base variety graph includes nodes indicating that the parents of the base variety are "variety ba1" and "variety ba2," as well as a node indicating that the base variety is "high-yielding." Such a base variety graph can be generated, for example, from various information recorded in a database that collects information on breeding.

[0041] For example, by accepting input of a base variety as a user request, information indicating the parent varieties of the base variety and the traits of the base variety can be extracted from the database described above, and a base variety graph can be generated. At this time, a request for traits desired for a new variety based on the base variety may also be accepted. Furthermore, the user may select a base variety from among the existing varieties shown in the existing variety graph. In this case, the existing variety graph of the selected existing variety can be used as the base variety graph.

[0042] By learning the existing variety graph as described above, it becomes possible to perform link prediction to determine which varieties are suitable for crossbreeding. In other words, the plant breeding support method according to this exemplary embodiment predicts the existing variety to be crossbreeded with the base variety by link prediction, and generates and outputs response information according to the predicted existing variety.

[0043] For example, in the example of Figure 4, link prediction is performed to determine which of the nodes representing various varieties included in the existing variety graph is likely to be connected to a node included in the base variety graph (more specifically, the "base variety" node) via a "good mating partner" link. Response information indicating that the predicted variety is an existing variety that should be bred with the base variety is then generated and output.

[0044] (Device configuration) The configuration of a plant breeding support device 2 according to a second exemplary embodiment of the present invention will be described with reference to Fig. 5. Fig. 5 is a block diagram showing the configuration of the plant breeding support device 2 according to this exemplary embodiment.

[0045] As shown in the figure, the breeding support device 2 includes a receiving unit 201, a graph generating unit 202, a learning unit 203, a link predicting unit 204, an evaluating unit 205, a generating unit 206, a basis generating unit 207, and an output unit 208.

[0046] In addition to these components, the breeding support device 2 may also include an input device that accepts user input operations, an output device that outputs data from the breeding support device 2, and a communication device that enables the breeding support device 2 to communicate with other devices. The output mode of the output device is arbitrary, and may be, for example, a display output or an audio output.

[0047] The reception unit 201 receives requests related to breeding. The requests may include, for example, information indicating the traits required for the new variety to be developed, the base variety on which the new variety will be based, and the like.

[0048] The graph generation unit 202 generates a base variety graph that represents the base variety in a graph based on information about the base variety that the user wants to develop as the basis for a new variety. Specifically, the graph generation unit 202 generates a base variety graph that represents the base variety as a node indicating the base variety, a node indicating the parent of the base variety, a node indicating the traits of the base variety, and edges that represent the relationships between the nodes. Note that the information about the base variety may be included in the request received by the receiving unit 201, or may be obtained from a database that accumulates information about breeding.

[0049] The learning unit 203 learns the relationships between the nodes included in the existing variety graph, in other words, the relationships between the properties of the existing varieties and the breeding process of the existing varieties, based on various information about the existing varieties, and generates a learned existing variety graph. The existing variety graph can also be said to be a graph that has been trained based on crossbreeding cases between existing varieties. Unless otherwise specified, the existing variety graph refers to one that has been trained by the learning unit 203. The trained existing variety graph may also be loaded into the breeding support device 2, in which case the learning unit 203 may be omitted.

[0050] The link prediction unit 204 uses a base variety graph including a plurality of nodes related to a base variety, which is one of the cross-parents of the new variety to be bred, and the above-mentioned existing variety graph to: base Variety graph and the above existing Link prediction is used to predict the relationships between nodes that are not connected by links in the variety graph, and existing varieties that should be crossed with the base variety are predicted.

[0051] The evaluation unit 205 evaluates the suitability of the mating candidate as a mating partner with the base variety based on the existing variety predicted by the link prediction unit 204, i.e., the nodes included in the existing variety graph of the mating candidate. The evaluation method will be described later.

[0052] The generation unit 206 generates response information including information about existing varieties that are candidates for breeding to develop a new variety, based on a trained model that has learned the relationships between the properties of existing varieties and the breeding processes of existing varieties, and on the request received by the reception unit 201. More specifically, the generation unit 206 generates response information according to the existing varieties (in other words, breeding candidates) predicted by the link prediction unit 204. As described above, the link prediction unit 204 performs link prediction using the existing variety graph, and therefore the generation unit 206 generates response information based on the results of link prediction by the link prediction unit 204, thereby generating response information based on the trained model.

[0053] The basis generating unit 207 generates basis information indicating the validity of the response information generated by the generating unit 206. The method of generating basis information will be described later.

[0054] The output unit 208 outputs various information generated by the breeding support device 2. For example, the output unit 208 outputs response information generated by the generation unit 206 and evidence information indicated by the evidence generation unit 207. The output destination of the information is arbitrary, and for example, if the breeding support device 2 is equipped with an output device as described above, the information may be output to that output device. Alternatively, for example, the information may be output to an output device external to the breeding support device 2.

[0055] As described above, the trained model used by the plant breeding support device 2 may be an existing variety graph, which includes multiple nodes related to existing varieties and links indicating the relationships between the nodes and is a trained graph of the relationships between the nodes. This configuration makes it possible to generate and output appropriate response information regarding existing varieties that are candidates for breeding, taking into consideration the interrelationships between the results of breeding existing varieties, their parents, their parents, traits, etc.

[0056] As described above, the plant breeding support device 2 uses a base variety graph including a plurality of nodes related to a base variety, which is one of the breeding parents of a new variety to be bred, and an existing variety graph in which examples of breeding between existing varieties have been learned, to: base The system may include a link prediction unit 204 that predicts existing varieties to be crossed with a base variety by link prediction for predicting the relationship between nodes not connected by links in the variety graph and the existing variety graph. The generation unit 206 may then generate response information according to the existing variety predicted by the link prediction unit 204. This makes it possible to recommend existing varieties that are considered appropriate based on past crossbreeding cases to the user as crossbreeding candidates.

[0057] (About link prediction) As described above, the link prediction unit 204 predicts existing varieties to be bred with the base variety by link prediction using the base variety graph and the existing variety graph. For example, as shown in the example of Figure 4, the link prediction unit 204 may predict existing variety nodes connected by a link between the "base variety" node and the "suitable breeding partner." The existing varieties predicted by the link prediction unit 204 become breeding candidates for the base variety.

[0058] The link prediction unit 204 may also predict that an existing variety that meets specified conditions is an existing variety to be crossed with a base variety. The conditions may be specified in advance or by the user. In the latter case, the receiving unit 201 may receive input of the conditions as a request.

[0059] For example, the link prediction unit 204 may predict that an existing variety that satisfies at least one of the following conditions should be bred with the base variety: (1) the variety obtained by crossing with an existing variety that has a specified trait maintains the specified trait, and (2) the variety has the trait that is desired to be added to the base variety.

[0060] When an existing variety that satisfies the above condition (1) is predicted as the existing variety to be crossed with the base variety, it is possible to predict the existing variety to be crossed with the base variety as a variety that is unlikely to lose a predetermined trait when crossed with the base variety.Furthermore, when an existing variety that satisfies the above condition (2) is predicted as the existing variety to be crossed with the base variety, it is possible to predict the existing variety to be crossed with the base variety as a variety that is likely to express a trait to be added to the base variety when crossed with the base variety.

[0061] (Evaluation of mating candidates) As described above, the evaluation unit 205 evaluates the suitability of the mating candidate as a mating partner with the base variety based on the nodes included in the existing variety graph of the mating candidate, i.e., the existing variety predicted by the link prediction unit 204. The evaluation by the evaluation unit 205 is described below.

[0062] Each node included in the existing variety graph of an existing variety predicted to be crossed with a base variety may indicate factors that influence the breeding of a new variety. For example, suppose the existing variety graph includes nodes and links indicating the number of crosses required to create the desired variety by crossing the existing variety as a parent. In this case, if the number of crosses indicated by those nodes and links is small, it is possible that the desired new variety can be created with a small number of crosses when the existing variety is crossed with the base variety.

[0063] Therefore, with the above configuration, the link prediction unit 204 evaluates the suitability of an existing variety as a breeding partner with a base variety based on the nodes included in the existing variety graph of the existing variety predicted by the link prediction unit 204. The evaluation method can be determined in advance based on the target nodes, etc. Then, the user can decide whether to use the predicted existing variety as a breeding partner with the base variety based on this evaluation. This can contribute to the appropriate selection of a variety to be bred with the base variety.

[0064] Various evaluation criteria can be applied. For example, evaluation may be based on the degree of suitability for the request. For example, assume that a trait desired in a new variety is included in the request. In this case, the evaluation unit 205 may evaluate an existing variety corresponding to an existing variety graph including a node indicating the trait higher than an existing variety corresponding to an existing variety graph not including a node indicating the trait.

[0065] The evaluation unit 205 may also express the evaluation results as a numerical value. In this exemplary embodiment, an example will be described in which the evaluation unit 205 calculates a recommendation level, which is a numerical value indicating the suitability of a breeding candidate as a mating partner with a base breed. In this case, for example, if a rule is established in advance regarding the relationship between the recommendation level and the nodes included in the existing breed graph of the breeding candidate, the evaluation unit 205 can calculate the recommendation level of each breeding candidate in accordance with the rule.

[0066] (Method of generating evidence information) As described above, the evidence generation unit 207 generates evidence information indicating the validity of the response information generated by the generation unit 206. For example, if the request accepted by the acceptance unit 201 includes information indicating traits desired for a new variety, the evidence generation unit 207 may generate evidence information including at least one of information regarding the breeding of the existing varieties to be used as breeding candidates and information regarding the traits of the existing varieties to be used as breeding candidates. This allows the user to refer to the response information in light of the evidence information and accurately determine the validity of the response information.

[0067] Information regarding the breeding of existing varieties that are candidates for breeding includes, for example, the parents of the existing varieties that are candidates for breeding, the traits of the parents, traits that were deteriorated or added during breeding to create the existing variety, the number of breedings, etc. Information regarding the traits of the varieties that are candidates for breeding includes, for example, traits that the varieties that are candidates for breeding have, or traits that the varieties that are candidates for breeding do not have, etc.

[0068] Various methods can be applied to generate evidence information. For example, assume that the request received by the receiving unit 201 includes information indicating traits desired for a new variety. In this case, the evidence generation unit 207 may check the presence or absence of the trait in the mating candidate and its parents, and if the presence of the trait is confirmed, generate evidence information indicating a variety that has the trait. Furthermore, if the presence of the trait is not confirmed in the parent generation, the evidence generation unit 207 may search back in the lineage until the trait is confirmed. Such confirmation of traits by tracing the lineage can be easily done using a graph of existing varieties.

[0069] (Generating evidence for link prediction results) The basis generating unit 207 can also generate basis information by analyzing the base product type graph and the existing product type graph. A method for generating basis information by analyzing the base product type graph and the existing product type graph will be described below.

[0070] For example, the basis generator 207 may mine one or more rules from the base variety graph and the existing variety graph using PCA (Principal Component Analysis) reliability based on the OWA (Open-world assumption).The basis generator 207 may then generate basis information using the mined one or more rules.For example, the method described in the following document may be applied to rule mining.

[0071] Luis Galarraga et. al, “Fast rule mining in ontological knowledge bases with AMIE +”, The VLDB Journal(2015)24:707-730 As an example, the rule to be processed by the basis generating unit 207 is expressed as follows, using Head r(x, y) and Body { B1 , . . . , Bn}:

number

number

[0072] The basis generating unit 207 sets the following as the conditions for the mining process: Connected: All values ​​(variables, entities) in the rule are shared between different atoms. Closed: All variables in the rule must appear at least twice. Not reflexive: Rules containing reflexive atoms such as r(x, x) are not mined. Mining processing is carried out under the following conditions.

[0073] Furthermore, the basis generating unit 207

number

number

[0074] For example, suppose that the basis generation unit 207 has mined a rule that "the breeding results will be good" when breeding varieties that satisfy the condition "there is no loss of desirable traits in the parent generation of the breed to be bred." In this case, when the link prediction unit 204 predicts an existing variety to be bred with the base variety, the basis generation unit 207 can generate basis information indicating that "there is no loss of desirable traits in the parent generation of the breed to be bred" as the basis for this prediction.

[0075] (Processing flow) The flow of the process (breeding support method) executed by the breeding support device 2 will be described with reference to Fig. 6. Fig. 6 is a flowchart showing the flow of the process executed by the breeding support device 2.

[0076] In S201, the reception unit 201 receives a request related to breeding. For example, in S201, a request indicating the traits required for a new variety to be created, a base variety, etc. is received. Next, in S202, the graph generation unit 202 generates a base variety graph based on the information input in S201.

[0077] In S203, the link prediction unit 204 predicts an existing variety to be bred with the base variety by link prediction using the base variety graph and the existing variety graph generated in S202, and determines this as a breeding candidate. The link prediction unit 204 may determine multiple breeding candidates. In this case, the basis generation unit 207 may generate basis information indicating the basis for the prediction result of the link prediction unit 204 by analyzing the base variety graph and the existing variety graph.

[0078] In S204, the evaluation unit 205 evaluates the suitability of the mating candidate as a mating partner with the base variety based on the nodes included in the existing variety graph of the mating candidate determined in S203. For example, the evaluation unit 205 may calculate a recommendation level indicating the suitability of the mating candidate as a mating partner with the base variety from the nodes included in the existing variety graph of the mating candidate. Note that if multiple mating candidates are determined in S203, the evaluation unit 205 evaluates each of the determined mating candidates.

[0079] In S205, the generation unit 206 generates response information based on the mating candidates determined in S203 and the request received in S201. As described above, the existing variety graph is a trained model that has learned the relationship between the properties of existing varieties and the breeding process of existing varieties. The mating candidates are determined by link prediction using the existing variety graph. Therefore, in S205, the response information is generated based on the trained model that has learned the relationship between the properties of existing varieties and the breeding process of existing varieties and the request received in S201.

[0080] For example, the generation unit 206 may generate response information indicating mating candidates determined in S203 that have an evaluation result of S204 that ranks up to a predetermined rank. Also, for example, the generation unit 206 may generate response information indicating mating candidates determined in S203 that match the request accepted in S201. Additionally, for example, the generation unit 206 may generate response information indicating the mating candidates determined in S203 and the evaluation result of S204.

[0081] In S206, the evidence generation unit 207 generates evidence information indicating the validity of the response information generated in S205. For example, the evidence generation unit 207 may refer to an existing variety graph to identify at least one of information regarding the breeding of the mating candidate and information regarding the traits of the mating candidate, and generate evidence information indicating the identified information.

[0082] In S207, the output unit 208 outputs the response information generated in S206. At this time, the output unit 208 may also output the basis information generated in S206. This completes the processing in FIG. 6.

[0083] (Example of response information) In S207, response information such as that shown in Fig. 7 may be output. Fig. 7 is a diagram showing an example of the response information. The response information shown in Fig. 7 includes six items: "Mating candidate," "Parents," "Number of matings," "Traits," "Maintenance of good traits of parent generation," and "Recommendation level."

[0084] "Breeding candidates" are predicted by the link prediction unit 204. In the example of Figure 7, varieties A to C are included in the breeding candidates. "Parent" indicates the parent variety of the breeding candidate. "Number of breedings" indicates the number of breedings required to produce the breeding candidate. "Traits" indicates the traits of the breeding candidate. "Maintenance of good traits from parent generation" indicates whether the breeding candidate maintains the good traits possessed by the parent generation (○: maintained, ×: unable to maintain). "Recommendation level" indicates the evaluation result of the evaluation unit 205 for the breeding candidate.

[0085] Among these items, "parents," "number of crosses," "traits," and "maintenance of good traits of parent generations" can be identified from the existing variety graph. The basis generating unit 207 may generate such information as basis information.

[0086] The "recommendation level" may be calculated by the evaluation unit 205 based on various information identified from the existing variety graph. For example, when evaluating based on the degree of suitability for a request, the evaluation unit 205 may calculate the recommendation level so that the recommendation level of a mating candidate corresponding to an existing variety graph including a node indicating a requested trait is higher than the recommendation level of a mating candidate corresponding to an existing variety graph that does not include a node indicating that trait. The evaluation unit 205 may also calculate the recommendation level taking into consideration the traits and number of matings of the mating candidate and its parents shown in the existing variety graph.

[0087] In the example of Figure 7, the recommendation levels for varieties A to C are 15, 5, and 0, respectively. For example, if rules are determined in advance, such as a recommendation level of +5 if the number of breedings is below a threshold, a recommendation level of +5 if one requested trait is possessed, and a recommendation level of +5 if the good traits of the parent generation are maintained, the evaluation unit 205 can calculate the recommendation level for each breeding candidate according to those rules.

[0088] Exemplary Embodiment 3 (overview) 8 is a diagram illustrating an overview of a plant breeding support method according to this exemplary embodiment. In this exemplary embodiment, an example of supporting plant breeding using a new variety graph including multiple nodes related to a new variety to be bred and multiple existing variety graphs generated for each of multiple existing varieties will be described.

[0089] In the plant breeding support method according to this exemplary embodiment, information for identifying a new variety to be developed is received as a breeding request. For example, a request may be received specifying the parents of the new variety to be developed, i.e., a base variety and a hybrid variety to be crossed with the base variety.

[0090] Next, in the plant breeding support method according to this exemplary embodiment, a new variety graph is generated based on the request. In the example of Figure 8, the new variety graph is a graph in which the "new variety" node is connected to the "hybrid variety" and "base variety" nodes by links called "parent."

[0091] In the plant breeding support method according to this exemplary embodiment, existing varieties that have a predetermined relationship with the new variety are identified by link prediction using the new variety graph generated as described above and the existing variety graph generated for multiple existing varieties. The existing variety graph used is generated for multiple existing varieties and has already learned the predetermined relationships between the multiple existing varieties.

[0092] In the example of Figure 8, existing varieties similar to the new variety (hereinafter also referred to as similar varieties) are predicted by link prediction using an existing variety graph that has learned the similarity of existing varieties A to D. Note that in this example, learning is performed so that dissimilar varieties are not connected by "similar" links (dissimilar varieties are treated as negative examples), but "dissimilar" links may also be learned. Furthermore, existing varieties A to D include nodes and links that indicate their parents, traits, etc., but these are omitted from Figure 8.

[0093] The types of varieties similar to a new variety are useful information for developing new varieties, as they are candidates for crossbreeding to develop new varieties. Therefore, by generating and outputting response information indicating the existing varieties identified as described above, it is possible to effectively support the breeding of new varieties.

[0094] Furthermore, the similar varieties identified as described above can be evaluated, and breeding candidates can be determined based on the evaluation results. For example, suppose a request is received indicating that the trait of a new variety to be developed is "disease resistance." In this case, if the identified similar varieties have the trait "disease resistance," response information may be generated recommending a hybrid variety included in the new variety graph as a breeding candidate for developing the new variety. On the other hand, if the identified similar varieties do not have the trait "disease resistance," response information may be generated indicating that the hybrid variety included in the new variety graph is unsuitable as a breeding candidate for developing the new variety. In this case, the user can simply make a request specifying a different hybrid variety.

[0095] (Device configuration) The configuration of a plant breeding support device 3 according to a third exemplary embodiment of the present invention will be described with reference to Fig. 9. Fig. 9 is a block diagram showing the configuration of the plant breeding support device 3 according to this exemplary embodiment.

[0096] As shown in the figure, the breeding support device 3 includes a receiving unit 301, a graph generating unit 302, a link predicting unit 303, an evaluating unit 304, a generating unit 305, a basis generating unit 306, and an output unit 307. Similar to the breeding support device 2 of the exemplary embodiment 2, the breeding support device 3 may include a learning unit, an input device, an output device, a communication device, etc. in addition to these components.

[0097] The reception unit 301 receives a breeding request. This request includes information for identifying the new variety to be created. Specifically, the request includes information indicating the base variety that will be used as the basis for the new variety and the hybrid variety to be crossed with the base variety. The request may also include information indicating the traits desired for the new variety to be created, the user's breeding needs, such as the upper limit of the number of crosses.

[0098] The graph generation unit 302 generates a new variety graph based on the request. For example, the graph generation unit 302 may generate a new variety graph (see FIG. 8) in which a node indicating a new variety is connected to a node indicating a base variety and a node indicating a hybrid variety by a link indicating that the base variety and the hybrid variety are the parents of the new variety. Note that the new variety graph may also include nodes and links indicating the parents, traits, etc. of the base variety and the hybrid variety.

[0099] The link prediction unit 303 identifies existing varieties that have a predetermined relationship with the new variety by link prediction using a new variety graph containing multiple nodes related to the new variety to be bred and an existing variety graph generated for multiple existing varieties. The predetermined relationship may be a relationship of similarity as in the example of Figure 8, or may be some other relationship. For example, the link prediction unit 303 may identify existing varieties that are dissimilar to the new variety, or it may also identify existing varieties that belong to the same classification as the new variety or existing varieties that have traits in common with the new variety.

[0100] The evaluation unit 304 evaluates the suitability of the hybrid variety included in the new variety graph as a mating partner for the base variety based on the nodes included in the existing variety graph of the existing variety identified by the link prediction unit 303, i.e., the existing variety having a predetermined relationship with the new variety. For example, if the link prediction unit 303 identifies a similar variety, the evaluation unit 304 may evaluate the similar variety as being suitable if the existing variety graph of the similar variety includes a node indicating the requested trait, or as not being suitable if the existing variety graph does not include a node indicating the requested trait.

[0101] The generation unit 305 generates response information including information about existing varieties that are candidates for breeding to develop a new variety, based on a trained model that has learned the relationships between the properties of existing varieties and the breeding processes of existing varieties, and on the request received by the reception unit 301. More specifically, the generation unit 305 generates response information about existing varieties identified by the link prediction unit 303. As described above, the link prediction unit 303 performs link prediction using the existing variety graph, which is a trained model. Therefore, the generation unit 305 generates response information based on the results of link prediction by the link prediction unit 303, thereby generating response information based on the trained model.

[0102] The generation unit 305 may also determine the suitability of a hybrid variety based on the evaluation results of the evaluation unit 304 and generate response information according to the determination results. For example, suppose that a request is made for the traits of a new variety to be developed, and the link prediction unit 303 identifies varieties similar to the new variety. In this case, the evaluation unit 304 may evaluate the identified similar varieties based on whether they possess the requested traits. If the evaluation unit 304 determines that the hybrid variety possesses the requested traits, the generation unit 305 may generate response information recommending the hybrid variety included in the new variety graph as a breeding candidate for developing the new variety. On the other hand, if the evaluation unit 304 determines that the hybrid variety does not possess the requested traits, the generation unit 305 may generate response information indicating that the hybrid variety included in the new variety graph is unsuitable as a breeding candidate for developing the new variety.

[0103] The evidence generation unit 306 generates evidence information indicating the validity of the response information generated by the generation unit 305. Specifically, the evidence generation unit 306 generates evidence information including at least one of information regarding the breeding of the existing variety to be a breeding candidate and information regarding the traits of the existing variety to be a breeding candidate. For example, the evidence generation unit 306 may generate evidence information indicating the parents, number of breedings, traits, etc. of the existing variety identified based on the nodes and links included in the existing variety graph of the existing variety identified by the link prediction unit 303. Furthermore, the evidence generation unit 306 may generate evidence information regarding the results of the link prediction by the link prediction unit 303 by analyzing the new variety graph and the existing variety graph.

[0104] The output unit 307 outputs the response information etc. generated by the generation unit 305. As with the output unit 208 in the second exemplary embodiment, the destination of the information output is not particularly limited.

[0105] As described above, the plant breeding support device 3 includes a link prediction unit 303 that identifies existing varieties that have a predetermined relationship with a new variety by link prediction using a new variety graph containing multiple nodes related to the new variety to be bred and an existing variety graph generated for multiple existing varieties.The generation unit 305 then generates response information related to the existing varieties identified by the link prediction unit 303. Information related to existing varieties that have a predetermined relationship with the new variety to be bred is useful in breeding new varieties, so this configuration makes it possible to provide information useful for breeding new varieties.

[0106] (Processing flow) The flow of the process (breeding support method) executed by the breeding support device 3 will be described with reference to Fig. 10. Fig. 10 is a flowchart showing the flow of the process executed by the breeding support device 3.

[0107] In S301, the receiving unit 301 receives a breeding request. For example, the request may indicate the traits desired for the new variety to be created, the base variety on which the new variety will be based, and the hybrid variety to be crossed with the base variety. Next, in S302, the graph generating unit 302 generates a new variety graph based on the information input in S301.

[0108] In S303, the link prediction unit 303 predicts similar varieties, which are existing varieties similar to the new variety, by link prediction using the new variety graph and the existing variety graph generated in S302. In addition, accompanying the processing of S303, the basis generation unit 306 may generate basis information indicating the basis for the prediction result of the link prediction unit 303 by analyzing the new variety graph and the existing variety graph.

[0109] In S304, the evaluation unit 304 evaluates the suitability of the hybrid variety included in the new variety graph as a mating partner with the base variety based on the nodes included in the existing variety graph of the similar variety predicted in S303. For example, the evaluation unit 304 may evaluate the similar variety as being compatible if the existing variety graph of the similar variety includes a node indicating the requested trait, or as being incompatible if it does not.

[0110] In S305, the generation unit 305 determines breeding candidates based on the similar varieties predicted in S303, and then in S306, the generation unit 305 generates response information indicating the breeding candidates determined in S305. As described above, the new variety graph is generated based on a request, the existing variety graph is a trained model, and similar varieties are identified by link prediction using the existing variety graph. Therefore, in S306, the response information is generated based on the trained model and the request.

[0111] For example, if the hybrid varieties included in the new variety graph are evaluated as suitable in S304, the generation unit 305 may determine the hybrid varieties included in the new variety graph as hybrid candidates for developing a new variety in S305, and generate response information indicating these hybrid candidates in S306. On the other hand, if the hybrid varieties are evaluated as unsuitable in S304, the generation unit 305 may not determine any hybrid candidates in S305, and may generate response information in S306 indicating that the hybrid varieties included in the new variety graph are unsuitable as hybrid candidates for developing a new variety.

[0112] The process of S305 may be omitted. In this case, the generating unit 305 may generate response information indicating the similar varieties predicted in S303 and the evaluation results in S304.

[0113] In S307, the evidence generation unit 305 generates evidence information indicating the validity of the response information generated in S306. For example, the evidence generation unit 306 may generate evidence information indicating the parents, number of breedings, traits, etc. of the similar varieties predicted in S303.

[0114] In S308, the output unit 307 outputs the response information generated in S306. At this time, the output unit 307 may also output the basis information generated in S307. This completes the processing in FIG. 10.

[0115] (Identifying varieties similar to the base variety) The link prediction unit 303 may identify a similar variety to a base variety instead of identifying a similar variety to a new variety. This will be described with reference to Fig. 11. Fig. 11 is a diagram showing an overview of a plant breeding support method including a process of identifying a similar variety to a base variety.

[0116] 11, link prediction is performed for existing varieties similar to the base variety (hereinafter referred to as similar varieties of the base variety) using the base variety graph and existing variety graphs of existing varieties A to D. Note that the base variety graph is the same as that described in the second exemplary embodiment, and the existing variety graph is the same as the example in FIG.

[0117] The existing variety graph shows the type of crossbreeding that produced the similar variety to the base variety, the varieties that have been crossbreeded with the similar variety to the base variety, and the varieties that have been produced by such crossbreeding. Therefore, by generating and outputting response information that includes information about the similar variety to the base variety, breeding can be favorably supported. For example, response information may be generated that indicates information about the similar variety to the base variety, varieties that have been crossbreeded with the similar variety to the base variety, varieties that have a similar variety to the base variety as a parent, or varieties that have a similar variety to the base variety as a parent.

[0118] It is also possible to identify existing varieties that have a relationship with the base variety other than similarity. For example, it is possible to identify existing varieties that are dissimilar to the base variety, or existing varieties that belong to the same classification as the base variety or that share traits with the base variety.

[0119] In addition, similar varieties to the base variety identified as described above can be evaluated, and breeding candidates can be determined based on the evaluation results. For example, suppose a request is received requesting that a new variety be developed with a trait of "disease resistance." In this case, if a progeny variety obtained by crossing a similar variety to the base variety with an existing variety has the trait of "disease resistance," response information can be generated recommending the existing variety as a breeding candidate for developing a new variety.

[0120] (Processing flow) The flow of processing (breeding support method) of the breeding support device 3 when identifying a similar variety to a base variety will be described with reference to Fig. 12. Fig. 12 is a flowchart showing the flow of processing executed by the breeding support device 3 when identifying a similar variety to a base variety.

[0121] In S301A, the receiving unit 301 receives a breeding request. In S301A, for example, a request indicating the traits desired for a new variety to be created and the base variety on which the new variety will be based is received. Next, in S302A, the graph generating unit 302 generates a base variety graph based on the information input in S301A.

[0122] In S303A, the link prediction unit 303 predicts a similar variety to the base variety by link prediction using the base variety graph and the existing variety graph generated in S302A. The link prediction unit 303 may predict multiple similar varieties to the base variety. In addition, in conjunction with the processing of S303A, the basis generation unit 306 may generate basis information indicating the basis for the prediction result of the link prediction unit 303 by analyzing the base variety graph and the existing variety graph.

[0123] In S304A, the evaluation unit 304 evaluates the similar variety based on the existing variety graph of the similar variety to the base variety predicted in S303A. For example, the evaluation unit 304 may evaluate the similar variety based on the degree to which the offspring variety of the similar variety matches the request. For example, the evaluation unit 304 may give a higher evaluation to a offspring variety that has the requested trait than to a offspring variety that does not have the requested trait. If multiple similar varieties are predicted in S303A, the evaluation unit 304 evaluates each of the determined similar varieties. The evaluation unit 304 may also consider generations below the offspring of the similar variety (grandchild generations and beyond) when making the evaluation.

[0124] In S305A, the generation unit 305 determines breeding candidates based on the similar varieties of the base variety predicted in S303A. In the following S306A, the generation unit 305 generates response information indicating the breeding candidates determined in S305A. As described above, the base variety graph is generated based on a request, the existing variety graph is a trained model, and the similar varieties are identified by link prediction using the existing variety graph. Therefore, in S306A, the response information is generated based on the trained model and the request.

[0125] For example, the generation unit 305 may generate response information indicating similar varieties predicted in S303A that have an evaluation result in S304A up to a predetermined rank. Furthermore, for example, the generation unit 305 may generate response information indicating similar varieties predicted in S303A that exhibit traits that match the request accepted in S301A in their offspring. Furthermore, the generation unit 305 may generate response information indicating varieties that have been crossbred with a similar variety to the base variety, together with or instead of the similar variety to the base variety, and that exhibit traits that match the request in their offspring.

[0126] In S307A, the evidence generation unit 306 generates evidence information indicating the validity of the response information generated in S306A. For example, the evidence generation unit 306 may refer to the existing variety graph to identify at least one of information related to the breeding of the mating candidate determined in S305 and information related to the traits of the mating candidate, and generate evidence information indicating the identified information.

[0127] In S308A, the output unit 307 outputs the response information generated in S306A. At this time, the output unit 307 may also output the basis information generated in S307A. This ends the processing in FIG. 12.

[0128] As described above, the link prediction unit 303 may identify an existing variety that has a predetermined relationship with the base variety by link prediction using the base variety graph and the existing variety graph. The generation unit 305 may then generate response information related to the existing variety identified by the link prediction unit 303. Information related to an existing variety that has a predetermined relationship with a base variety that is one of the breeding parents of a new variety is useful in breeding new varieties, so the above configuration can provide information useful for breeding using the base variety.

[0129] Exemplary Embodiment 4 (overview) 13 is a diagram illustrating an overview of a plant breeding support method according to this exemplary embodiment. In this exemplary embodiment, an example is described in which a new variety graph including multiple nodes related to new varieties to be bred is updated while searching for breeding candidates for new varieties that meet a request.

[0130] In this exemplary embodiment, link prediction is performed using the new variety graph and the existing variety graph, as in exemplary embodiment 3. The new variety graph shown in the top left corner of Fig. 13 includes nodes and links indicating that the parents of the new variety are "variety x" and "base variety."

[0131] 13 includes nodes and links indicating that existing variety A has the trait of "disease resistance," existing variety B has the trait of "low-temperature tolerance," and existing variety C has the traits of "good taste" and "high yield." Note that other nodes and links are omitted from the illustration.

[0132] By learning the existing variety graphs for various existing varieties as described above, it becomes possible to link-predict which varieties are likely to have which traits. In other words, the plant breeding support method according to this exemplary embodiment generates a tentative new variety graph and performs link-prediction of the probability that a new variety shown in the new variety graph will have a requested trait.

[0133] For example, in the example in Figure 13, the probability that the "Disease Resistance" node will be connected to the "New Variety" node in the new variety graph shown on the top left is predicted to be 30% via a "Trait" link. This probability cannot be said to be sufficiently high.

[0134] Therefore, as shown in the bottom left of the figure, the node connected to the "new variety" node in the new variety graph by the "parent" link is changed from "variety x" to "variety y," and link prediction is performed again. As a result, the predicted probability that the "disease resistance" node will be connected to the "new variety" node by the "trait" link changes to 80%.

[0135] According to the plant breeding support method of this exemplary embodiment, from the results of the above processing, it is possible to recommend "variety y" as a candidate for breeding to impart the trait of disease resistance to a new variety.

[0136] (Device configuration) The configuration of a plant breeding support device 4 according to a fourth exemplary embodiment of the present invention will be described with reference to Fig. 14. Fig. 14 is a block diagram showing the configuration of the plant breeding support device 4 according to this exemplary embodiment.

[0137] As shown in the figure, the breeding support device 4 includes a receiving unit 401, a graph generating unit 402, a link predicting unit 403, a graph updating unit 404, a generating unit 405, a basis generating unit 406, and an output unit 407. Similar to the breeding support device 2 of the exemplary embodiment 2, the breeding support device 4 may include a learning unit, an input device, an output device, a communication device, etc. in addition to these components.

[0138] The reception unit 401 receives a breeding request. This request includes information for identifying the new variety to be developed. For example, the request may include information indicating a base variety that will be used as the basis for the new variety and a hybrid variety to be crossed with the base variety. The request may also include information indicating the traits desired for the new variety to be developed and information indicating the user's breeding needs, such as an upper limit on the number of crosses.

[0139] The graph generation unit 402 generates a new variety graph based on the request. For example, the graph generation unit 402 may generate a new variety graph (see FIG. 13 ) in which nodes representing the base variety and the hybrid variety are connected to a node representing the new variety by links indicating that they are the parents of the new variety. The new variety graph may also include nodes and links indicating the parents, traits, etc. of the base variety and the hybrid variety.

[0140] The link prediction unit 403 calculates the probability that a node showing a predetermined trait will link to a node included in the new variety graph by link prediction using the new variety graph generated by the graph generation unit 402 and the learned existing variety graph. The predetermined trait is specified based on a request. For example, when a trait required for a new variety to be created is requested, the link prediction unit 403 calculates the probability that a node showing the trait will link to a node included in the new variety graph (for example, the "new variety" node in the example of FIG. 13).

[0141] The graph update unit 404 updates the new variety graph. Specifically, the graph update unit 404 performs processing to replace the node indicating the parent of the new variety, which is included in the new variety graph, with a node of another variety.

[0142] The new variety graph may be updated according to user input or automatically. In the former case, the graph update unit 404 may cause the output unit 407 to output a list of existing varieties extracted from the existing variety graph, and allow the user to select new breeding candidates from that list. In the latter case, the graph update unit 404 may select new breeding candidates from the existing varieties extracted from the existing variety graph. New breeding candidates may be selected from existing varieties whose offspring have individuals with the requested trait.

[0143] The generation unit 405 generates response information including information about existing varieties that are candidates for breeding to develop a new variety, based on a trained model that has learned the relationship between the properties of existing varieties and the breeding process of existing varieties, and on the request received by the reception unit 401. More specifically, the generation unit 405 generates the response information based on the probability calculated by the link prediction unit 403. A specific example of generating response information will be described later with reference to FIG. 15.

[0144] As described above, the link prediction unit 403 performs link prediction using the existing variety graph, which is a trained model, and the new variety graph generated based on the request. Therefore, the generation unit 405 generates response information based on the result of link prediction by the link prediction unit 403, thereby generating response information based on the trained model and the request.

[0145] The evidence generation unit 406 generates evidence information indicating the validity of the response information generated by the generation unit 405. Specifically, the evidence generation unit 406 generates evidence information including at least one of information regarding the breeding of the existing variety to be used as a breeding candidate and information regarding the traits of the existing variety to be used as a breeding candidate. The evidence generation unit 406 may also generate evidence information regarding the results of the link prediction by the link prediction unit 403 by analyzing the new variety graph and the existing variety graph.

[0146] The output unit 407 outputs the response information etc. generated by the generation unit 405. As with the output unit 208 in the second exemplary embodiment, the destination of the information output is not particularly limited.

[0147] As described above, the plant breeding support device 4 includes a link prediction unit 403 that calculates the probability that a node exhibiting a predetermined trait will be linked to a node included in the new variety graph by link prediction using the new variety graph and the existing variety graph. The generation unit 405 then generates response information based on the probability calculated by the link prediction unit 403. The probability that a node exhibiting a predetermined trait will be linked to the new variety graph indicates the possibility that the new variety will have the predetermined trait, and the response information generated based on this probability is useful in breeding new varieties. Therefore, the above configuration can provide information useful for breeding new varieties with desired traits.

[0148] (Processing flow) The flow of the process (breeding support method) executed by the breeding support device 4 will be described with reference to Fig. 15. Fig. 15 is a flowchart showing the flow of the process executed by the breeding support device 4.

[0149] In S401, the reception unit 401 receives a request regarding breeding. In S401, for example, a request indicating a base variety to be used as the base for a new variety, a hybrid variety to be crossed with the base variety, and traits desired for the new variety to be produced is received.

[0150] In S402, the graph generation unit 402 generates a new variety graph based on the information input in S401. For example, if input of a base variety and a hybrid variety is received in S401, the graph generation unit 402 may generate a new variety graph including nodes and links indicating that those varieties are the parents of the new variety.

[0151] In S403, the link prediction unit 403 calculates the probability that a node showing a trait that matches the request received in S401 will be linked to a node included in the new variety graph generated in S402. As described above, this probability is calculated by link prediction using the learned existing variety graph and the new variety graph. In addition to the processing of S403, the basis generation unit 406 may generate basis information indicating the basis for the calculation result of the link prediction unit 403 by analyzing the new variety graph and the existing variety graph.

[0152] In S404, the graph update unit 404 determines whether the probability calculated in S403 is equal to or greater than a threshold value. If it is determined that the probability is equal to or greater than the threshold value (YES in S404), the process proceeds to S406, and if it is determined that the probability is less than the threshold value (NO in S404), the process proceeds to S405.

[0153] If the request received in S401 specifies multiple traits, predictions are made for each trait in S403, and if the probability for all traits is equal to or greater than a threshold, a YES result is made in S404, and if even one trait is below the threshold, a NO result is made. This makes it possible to estimate a hybrid variety that can produce a new variety possessing all the requested traits.

[0154] In S405, the graph update unit 404 updates the new variety graph. Specifically, the graph update unit 404 replaces the node of the hybrid variety included in the current new variety graph with a node of another variety. Note that the graph update unit 404 may also change the base variety. As described above, the update contents may be determined according to user input, or may be determined by the graph update unit 404.

[0155] Once the new variety graph has been updated, the process returns to S403, and the probability is calculated again. That is, in the process of Fig. 15, the calculation of the probability in S403 and the updating of the new variety graph in S405 are repeated until a YES determination is made in S404.

[0156] In S406, the generation unit 405 estimates the breeding parents to be used to create the new variety that matches the request received in S401, and generates response information indicating the estimated breeding parents. Specifically, the generation unit 405 estimates that the breeding parents shown in the new variety graph when the determination in S404 is YES are the breeding parents to be used to create the new variety that matches the request, and generates response information indicating the breeding parents. Note that the estimated breeding parents may be only one (the one other than the base variety) or both.

[0157] In S407, the evidence generation unit 406 generates evidence information indicating the validity of the response information generated in S406. Specifically, the evidence generation unit 406 generates evidence information including at least one of information regarding the breeding of the existing breeding candidate and information regarding the traits of the existing breeding candidate.

[0158] In S408, the output unit 407 outputs the response information generated in S406. At this time, the output unit 407 may also output the basis information generated in S407. This ends the processing in FIG. 15.

[0159] The breeding support method according to this exemplary embodiment can also be used for breeding over multiple generations. In this case, after determining breeding candidates as described above, the new variety graph can be updated to include a node representing the offspring produced by the breeding of these varieties, and breeding candidates for the offspring can then be determined. By repeating this process until a new variety with the desired traits is found, it becomes possible to support the creation of new varieties that are difficult to produce in a single breeding.

[0160] Exemplary Embodiment 5 (overview) 16 is a diagram illustrating an overview of a breeding support method according to this exemplary embodiment. In this exemplary embodiment, an example is described in which the suitability of an existing variety as a breeding parent is evaluated using an existing variety graph that includes at least a breeding parent node indicating the breeding parent of the existing variety, and response information is generated based on the evaluation results.

[0161] The existing variety graph for variety A shown in Figure 16 indicates that the parents of variety A are varieties a1 and a2, the parents of variety a1 are varieties a11 and a12, and the parents of variety a2 are varieties a21 and a22. The nodes that indicate the breeding parents of such existing varieties are the breeding parent nodes described above.

[0162] In the plant breeding support method according to this exemplary embodiment, the suitability of an existing variety as a breeding parent is evaluated based on such breeding parent nodes. For example, when evaluating the suitability of existing variety A shown in Figure 16 as a breeding parent to be mated with a base variety, the evaluation is performed based on the breeding parent nodes of existing variety A (varieties a1, a2, a11, a12, a21, and a22).

[0163] Specifically, although not shown in Figure 16, each mating parent node contains nodes and links that indicate the traits, etc. of that variety, so existing variety A can be evaluated based on these nodes and links.

[0164] The evaluation criteria may be predetermined. For example, evaluation may be based on suitability to a user's breeding request. For example, suppose a trait desired for a new variety is requested. In this case, if a node indicating the trait is linked to a breeding parent node, it can be said that the trait will likely be inherited by offspring. Therefore, an existing variety in which a node indicating the requested trait is linked to a breeding parent node may be rated higher than an existing variety that does not have such a link to a breeding parent node.

[0165] Furthermore, if a node indicating the inheritance of favorable traits from the parent generation is linked to a breeding parent node, it can be said that the favorable traits of the parent generation are likely to be inherited by the offspring generation. Therefore, an existing variety that has a node indicating the inheritance of favorable traits from the parent generation linked to a breeding parent node may be rated higher than an existing variety that does not have such a link to a breeding parent node.

[0166] By evaluating each existing variety in the above manner, it is possible to estimate highly rated existing varieties, i.e., existing varieties that should be used as breeding parents for new varieties, and response information indicating this estimation result can be generated and output. In this way, the plant breeding support method according to this exemplary embodiment can generate response information useful for breeding new varieties, taking into account information about the breeding parents of existing varieties.

[0167] In particular, the breeding support method according to this exemplary embodiment uses an existing variety graph to trace not only the parents of existing varieties that are candidates for breeding, but also their parents and their parents, and so on, through multiple stages of breeding lines, and evaluates the breeding candidates taking these into consideration. This makes it possible to perform a valid evaluation of each existing variety and generate and output valid response information based on the valid evaluation results.

[0168] While Figure 16 shows an example of evaluating an existing variety to be crossed with a base variety, an existing variety to be used as a base variety can also be evaluated in a similar manner. This allows identifying existing varieties that are suitable as base varieties.

[0169] (Device configuration) The configuration of a plant breeding support device 5 according to a fifth exemplary embodiment of the present invention will be described with reference to Fig. 17. Fig. 17 is a block diagram showing the configuration of the plant breeding support device 5 according to this exemplary embodiment.

[0170] As shown in the figure, the breeding support device 5 includes a receiving unit 501, an evaluating unit 502, a generating unit 503, and an output unit 504. Similar to the breeding support device 2 of the exemplary embodiment 2, the breeding support device 5 may include a learning unit 203, a basis generating unit 207, an input device, an output device, a communication device, and the like in addition to these components.

[0171] The reception unit 501 receives a breeding request. This request may include the user's wishes regarding breeding, such as traits desired for a new variety to be created. This request may also include information indicating a base variety.

[0172] The evaluation unit 502 evaluates the suitability of each existing variety as a breeding parent for use in breeding candidates to create a new variety, based on the breeding parent node of the existing variety. As explained with reference to Figure 16, various evaluation methods can be applied, and the evaluation may be performed taking the request into consideration or without taking the request into consideration.

[0173] The evaluation unit 502 may also perform the evaluation by taking into consideration factors other than the breeding parent node (for example, various information related to the base variety). For example, the evaluation unit 502 may give a higher rating to an existing variety that has a breeding parent with a trait not found in the base variety than to an existing variety that does not have a breeding parent with such a trait. The evaluation unit 502 may perform the evaluation based on multiple criteria and combine the evaluation results based on each criterion to arrive at a final evaluation result.

[0174] The generation unit 503 generates response information based on the evaluation results of the evaluation unit 502. For example, the generation unit 503 may generate response information indicating the existing varieties evaluated by the evaluation unit 502 and the evaluation results of the existing varieties. Furthermore, for example, the generation unit 503 may infer that a predetermined number of existing varieties with the highest evaluation results by the evaluation unit 502 are existing varieties that should be used as parents for breeding of the new variety, and generate response information indicating the inferred result.

[0175] When the evaluation unit 502 performs the evaluation without taking the request into consideration, the generation unit 503 generates response information while taking the request into consideration. For example, the generation unit 503 may estimate that an existing variety with the highest evaluation results by the evaluation unit 502 that has the highest suitability for the request is the existing variety that should be used as a breeding parent for the new variety. The generation unit 503 may then generate response information indicating the estimation result. Suitability for the request may be determined based on nodes directly or indirectly linked to the existing variety.

[0176] The output unit 504 outputs the response information etc. generated by the generation unit 503. As with the output unit 208 in the second exemplary embodiment, the destination of the information output is not particularly limited.

[0177] As described above, the plant breeding support device 5 uses an existing variety graph including at least a breeding parent node indicating the breeding parent of the existing variety as a trained model that has learned the relationship between the properties of the existing variety and the breeding process of the existing variety. The plant breeding support device 5 further includes an evaluation unit 502 that evaluates the suitability of the existing variety as a breeding parent based on at least the breeding parent node, and a generation unit 503 that generates response information based on the evaluation result of the evaluation unit 502. This makes it possible to generate response information useful for breeding new varieties, taking into account information about the breeding parent of the existing variety.

[0178] (Processing flow) The flow of the process (breeding support method) executed by the breeding support device 5 will be described with reference to Fig. 18. Fig. 18 is a flowchart showing the flow of the process executed by the breeding support device 5.

[0179] In S501, the receiving unit 501 receives a request regarding breeding. In S501, for example, a request including a user's desire regarding breeding, such as traits required for a new variety to be created, is received.

[0180] In S502, the evaluation unit 502 evaluates the breeding candidates. More specifically, for each existing variety that is a breeding candidate to be used to create a new variety, the evaluation unit 502 evaluates the suitability of the existing variety as a breeding parent based on the breeding parent node of the existing variety.

[0181] In S503, the generating unit 503 generates response information based on the evaluation results of S502. For example, the generating unit 503 may infer that a predetermined number of existing varieties with the highest evaluation results of S502 are existing varieties that should be used as breeding parents for the new variety, and generate response information indicating this inference.

[0182] If the breeding support device 5 is equipped with the evidence generation unit 207, the evidence generation unit 207 generates evidence information indicating the validity of the response information generated in S503. In this case, the evidence generation unit 207 generates evidence information including at least one of information on the breeding of the existing breeding candidate and information on the traits of the existing breeding candidate.

[0183] In S504, the output unit 504 outputs the response information generated in S503. If the basis information has been generated as described above, the output unit 504 may also output the basis information. This completes the processing in FIG. 18.

[0184] [Modification] As explained in the fourth exemplary embodiment, by using the new variety graph and the existing variety graph, it is possible to predict the traits of a new variety by link prediction. Furthermore, trait prediction of a new variety can also be performed by methods other than link prediction. This will be explained with reference to FIG. 19. FIG. 19 is a diagram illustrating an example of predicting the traits of a new variety based on feature quantities calculated from the new variety graph and the existing variety graph. FIG. 19 shows the existing variety graphs of existing varieties A to C and the new variety graph of the new variety. Note that of the nodes and links included in these graphs, only those indicating that the parents of the new variety are hybrid varieties of the base variety are not shown in the illustration.

[0185] Here, the feature values ​​for each existing variety can be calculated by multiplying the feature values ​​of each node included in the existing variety graph by a weight corresponding to the link connected to that node and adding the results together. Therefore, if learning is performed to update the weights so that the calculated feature values ​​correspond to the traits of the existing variety, it will be possible to predict the traits of a new variety from the feature values ​​of the new variety graph calculated by applying the weights.

[0186] For example, in the example of Figure 19, the feature calculated from the existing variety graph of existing variety A, which is known to have good eating taste, is trained to fall within a range in the feature space corresponding to the trait "good eating taste." Also, the feature calculated from the existing variety graph of existing variety B, which is known to have disease resistance, is trained to fall within a range in the feature space corresponding to the trait "disease resistance." Similarly, the feature calculated from the existing variety graph of existing variety C, which is known to have high temperature resistance, is trained to fall within a range in the feature space corresponding to the trait "high temperature resistance."

[0187] In this case, as shown in the figure, if the feature values ​​calculated from the new variety graph are within the ranges corresponding to the traits "good eating taste" and "disease resistance," it can be predicted that the new variety has the traits "good eating taste" and "disease resistance." Such a trait prediction method can be applied as an alternative to the trait prediction method in the above-described exemplary embodiment.

[0188] [Software implementation example] Some or all of the functions of the plant breeding support devices 1 to 5 (hereinafter referred to as the devices) may be realized by hardware such as an integrated circuit (IC chip), or by software.

[0189] In the latter case, the device is realized by, for example, a computer that executes instructions of a program, which is software that realizes each function. An example of such a computer (hereinafter referred to as computer C) is shown in FIG. 20. Computer C includes at least one processor C1 and at least one memory C2. Memory C2 stores a program (breeding support program) P for operating computer C as the device. In computer C, processor C1 reads and executes program P from memory C2, thereby realizing each function of the device.

[0190] The processor C1 may be, for example, a central processing unit (CPU), a graphics processing unit (GPU), a digital signal processor (DSP), a micro processing unit (MPU), a floating point number processing unit (FPU), a physics processing unit (PPU), a microcontroller, or a combination thereof. The memory C2 may be, for example, a flash memory, a hard disk drive (HDD), a solid state drive (SSD), or a combination thereof.

[0191] The computer C may further include a RAM (Random Access Memory) for expanding the program P during execution and for temporarily storing various data. The computer C may also include a communication interface for transmitting and receiving data to and from other devices. The computer C may also include an input / output interface for connecting input / output devices such as a keyboard, mouse, display, and printer.

[0192] Furthermore, the program P can be recorded on a non-transitory tangible recording medium M that can be read by the computer C. Such a recording medium M can be, for example, a tape, a disk, a card, a semiconductor memory, or a programmable logic circuit. The computer C can acquire the program P via such a recording medium M. The program P can also be transmitted via a transmission medium. Such a transmission medium can be, for example, a communication network or broadcast waves. The computer C can also acquire the program P via such a transmission medium.

[0193] [Appendix 1] The present invention is not limited to the above-described embodiments, and various modifications are possible within the scope of the claims. For example, embodiments obtained by appropriately combining the technical means disclosed in the above-described embodiments are also included in the technical scope of the present invention.

[0194] [Appendix 2] Some or all of the above-described embodiments can also be described as follows: However, the present invention is not limited to the following described aspects.

[0195] (Appendix 1) A breeding support device comprising: a receiving means for receiving a request for breeding, a trained model that has learned the relationship between the properties of existing varieties and the breeding process of the existing varieties, a generating means for generating response information including information on existing varieties that are candidates for breeding to develop a new variety based on the request, and an output means for outputting the response information. This configuration has the effect of providing effective support for breeding.

[0196] (Appendix 2) The request includes information indicating traits desired for the new variety, and the breeding support device further includes a basis generating unit configured to generate basis information including at least one of information on the breeding of the existing candidate varieties and information on the traits of the existing candidate varieties. This allows a user to refer to the response information in light of the basis information, thereby enabling the user to accurately determine the validity of the response information.

[0197] (Appendix 3) The plant breeding support device according to claim 1 or 2, wherein the trained model is an existing variety graph, which includes a plurality of nodes related to the existing varieties and links indicating the relationships between the nodes, and is a trained graph of the relationships between the nodes. With this configuration, it is possible to generate and output appropriate response information regarding existing varieties that are candidates for breeding, taking into consideration the interrelationships between the results of breeding of existing varieties, their parents, their parents, traits, etc.

[0198] (Appendix 4) The plant breeding support device according to Appendix 3 further comprises a link prediction means for predicting an existing variety to be bred with a base variety by link prediction using a base variety graph including a plurality of nodes related to a base variety, which is one of the breeding parents of a new variety to be bred, and the existing variety graph, for predicting relationships between nodes that are not connected by links in the base variety graph and the existing variety graph, and the generation means generates the response information according to the existing variety predicted by the link prediction means. This makes it possible to recommend to the user existing varieties that are deemed appropriate based on past breeding cases as breeding candidates.

[0199] (Appendix 5) The link prediction means predicts that an existing variety that satisfies at least one of the following conditions is an existing variety to be bred with the base variety: (1) a variety obtained by crossing with an existing variety having a specified trait maintains the specified trait; and (2) the variety has a trait that is desired to be added to the base variety.

[0200] When an existing variety that satisfies the above condition (1) is predicted as the existing variety to be crossed with the base variety, it is possible to predict the existing variety to be crossed with the base variety as a variety that is unlikely to lose a predetermined trait when crossed with the base variety.Furthermore, when an existing variety that satisfies the above condition (2) is predicted as the existing variety to be crossed with the base variety, it is possible to predict the existing variety to be crossed with the base variety as a variety that is likely to express a trait to be added to the base variety when crossed with the base variety.

[0201] (Appendix 6) The plant breeding support device according to claim 4, further comprising an evaluation means for evaluating the suitability of the existing variety as a mating partner with the base variety based on the nodes included in the existing variety graph of the existing variety predicted by the link prediction means, thereby contributing to the appropriate selection of a variety to be mated with the base variety.

[0202] (Appendix 7) The plant breeding support device according to claim 3 further comprises a link prediction means for identifying existing varieties that have a predetermined relationship with the new variety by link prediction using a new variety graph including a plurality of nodes related to the new variety to be bred and the existing variety graph generated for a plurality of the existing varieties, and the generation means generates the response information related to the existing varieties identified by the link prediction means. This configuration makes it possible to provide information useful for breeding new varieties.

[0203] (Appendix 8) The plant breeding support device according to claim 3, further comprising a link prediction means for identifying existing varieties that have a predetermined relationship with a base variety by link prediction using a base variety graph including a plurality of nodes relating to a base variety that is one of the breeding parents of a new variety to be bred and the existing variety graph generated for a plurality of the existing varieties, and the generation means for generating the response information relating to the existing varieties identified by the link prediction means. This can provide information useful for breeding using a base variety.

[0204] (Appendix 9) The plant breeding support device according to claim 3 further comprises a link prediction means for calculating the probability that a node exhibiting a predetermined trait will be linked to a node included in a new variety graph by link prediction using a new variety graph including a plurality of nodes related to a new variety to be bred and the existing variety graph, and the generation means generates the response information based on the probability calculated by the link prediction means. This configuration makes it possible to provide information useful for breeding new varieties having desired traits.

[0205] (Appendix 10) The plant breeding support device according to claim 1 or 2, wherein the trained model is an existing variety graph including at least a breeding parent node indicating a breeding parent of the existing variety, and the device comprises an evaluation means for evaluating the suitability of the existing variety as a breeding parent based on at least the breeding parent node, and the generation means generates the response information based on the evaluation result of the evaluation means. With this configuration, response information useful for breeding new varieties can be generated taking into account information on the breeding parents of the existing variety.

[0206] (Appendix 11) A method for supporting plant breeding in which a computer receives a request for breeding, generates response information including information on existing varieties that are candidates for breeding to develop a new variety based on the request using a trained model that has learned the properties of existing varieties and the relationships between the breeding processes of the existing varieties, and outputs the response information. This configuration has the effect of providing effective support for breeding.

[0207] (Appendix 12) A breeding support program that causes a computer to execute the following processes: accepting a request for breeding, generating response information containing information about existing varieties that are candidates for breeding to develop a new variety based on the request using a trained model that has learned the relationship between the properties of existing varieties and the breeding process of the existing varieties, and outputting the response information. This configuration provides the effect of providing optimal support for breeding.

[0208] [Appendix 3] Some or all of the above-described embodiments can also be expressed as follows. A plant breeding support device comprising at least one processor that performs the following processes: accepting a request for breeding; generating response information based on the request using a trained model that has learned the relationship between the properties of existing varieties and the breeding process of the existing varieties, including information about existing varieties that are candidates for breeding to develop a new variety; and outputting the response information.

[0209] The plant breeding support device may further include a memory that stores a program (breeding support program) for causing the processor to execute the following steps: accepting a request for breeding; generating response information including information about existing varieties that are candidates for breeding to develop a new variety based on the request using a trained model that has learned the relationship between the properties of existing varieties and the breeding processes of the existing varieties; and outputting the response information. The program may also be recorded on a computer-readable, non-transitory, tangible recording medium. [Explanation of symbols]

[0210] 1 Breed improvement support device: 11 reception unit, 12 generation unit, 13 output unit 2 Breed improvement support device: 201 reception unit, 204 link prediction unit, 205 evaluation unit, 206 generation unit, 207 evidence generation unit, 208 output unit 3 Breed improvement support device: 301 reception unit, 303 link prediction unit, 304 evaluation unit, 305 generation unit, 306 Basis generation unit, 307 Output unit 4 Breed improvement support device: 401 reception unit, 403 link prediction unit, 405 generation unit, 406 evidence generation unit, 407 Output Section 5 Breed improvement support device: 501 reception unit, 502 evaluation unit, 503 generation unit, 504 output unit

Claims

1. a receiving means for receiving requests regarding breeding; a generation means for generating response information including information on existing varieties that are candidates for breeding to develop a new variety, based on a trained model that has learned the relationship between the properties of existing varieties and the breeding process of the existing varieties, and the request; an output means for outputting the response information; Equipped with The trained model is an existing variety graph that includes a plurality of nodes related to the existing varieties and links indicating relationships between the nodes, and is a trained graph of relationships between the nodes; and the request includes information indicating traits desired for the new variety; a basis generating means for generating basis information including at least one of information on the breeding of the existing varieties to be the crossbreeding candidates and information on the traits of the existing varieties to be the crossbreeding candidates; A breeding support device comprising:

2. A reception means for receiving requests regarding breeding; a generation means for generating response information including information on existing varieties that are candidates for breeding to develop a new variety, based on a trained model that has learned the relationship between the properties of existing varieties and the breeding process of the existing varieties, and the request; an output means for outputting the response information; Equipped with The trained model is an existing variety graph that includes a plurality of nodes related to the existing varieties and links indicating relationships between the nodes, and is a trained graph of relationships between the nodes; and a link prediction means for predicting an existing variety to be bred with a base variety by link prediction using a base variety graph including a plurality of nodes related to a base variety, which is one of the breeding parents of a new variety to be bred, and the existing variety graph, for predicting a relationship between nodes not connected by a link in the base variety graph and the existing variety graph; The generation means generates the response information according to the existing variety predicted by the link prediction means.

3. The link prediction means determines whether (1) a variety obtained by crossbreeding with an existing variety having a predetermined trait maintains the predetermined trait, and (2) whether the base 3. The plant breeding support device according to claim 2, wherein an existing variety that satisfies at least one of the conditions above is predicted as an existing variety to be crossed with the base variety.

4. 3. The breeding support device according to claim 2, further comprising an evaluation means for evaluating the suitability of the existing variety as a breeding partner for the base variety based on the nodes included in the existing variety graph of the existing variety predicted by the link prediction means.

5. A reception means for receiving requests regarding breeding; a generation means for generating response information including information on existing varieties that are candidates for breeding to develop a new variety, based on a trained model that has learned the relationship between the properties of existing varieties and the breeding process of the existing varieties, and the request; an output means for outputting the response information; Equipped with The trained model is an existing variety graph that includes a plurality of nodes related to the existing varieties and links indicating relationships between the nodes, and is a trained graph of relationships between the nodes; and a link prediction means for identifying existing varieties that have a predetermined relationship with the new variety by link prediction using a new variety graph including a plurality of nodes related to the new variety to be bred and the existing variety graph generated for a plurality of the existing varieties; The generation means generates the response information regarding the existing variety identified by the link prediction means.

6. A reception means for receiving requests regarding breeding; a generation means for generating response information including information on existing varieties that are candidates for breeding to develop a new variety, based on a trained model that has learned the relationship between the properties of existing varieties and the breeding process of the existing varieties, and the request; an output means for outputting the response information; Equipped with The trained model is an existing variety graph that includes a plurality of nodes related to the existing varieties and links indicating relationships between the nodes, and is a trained graph of relationships between the nodes; and a link prediction means for identifying existing varieties that have a predetermined relationship with a base variety by link prediction using a base variety graph including a plurality of nodes relating to a base variety that is one of the breeding parents of a new variety to be bred, and the existing variety graph generated for a plurality of the existing varieties; The generation means generates the response information regarding the existing variety identified by the link prediction means.

7. A reception means for receiving requests regarding breeding; a generation means for generating response information including information on existing varieties that are candidates for breeding to develop a new variety, based on a trained model that has learned the relationship between the properties of existing varieties and the breeding process of the existing varieties, and the request; an output means for outputting the response information; Equipped with The trained model is an existing variety graph that includes a plurality of nodes related to the existing varieties and links indicating relationships between the nodes, and is a trained graph of relationships between the nodes; and a link prediction means for calculating the probability that a node showing a predetermined trait will be linked to a node included in a new variety graph by link prediction using a new variety graph including a plurality of nodes related to a new variety to be bred and the existing variety graph; The generation means generates the response information based on the probability calculated by the link prediction means.

8. the trained model is an existing variety graph including at least a breeding parent node indicating a breeding parent of the existing variety, an evaluation means for evaluating the suitability of the existing variety as a breeding parent based on at least the breeding parent node; 2. The plant breeding support device according to claim 1, wherein said generating means generates said response information based on an evaluation result of said evaluating means.

9. The computer Accepting breeding requests, generating, based on the request, response information including information about existing varieties that are candidates for breeding to develop a new variety, using a trained model that has learned the properties of existing varieties and the relationship between the breeding processes of the existing varieties; generating evidence information including at least one of information on the breeding of the existing varieties to be the crossbreeding candidates and information on the traits of the existing varieties to be the crossbreeding candidates; outputting the response information and the basis information; The trained model is an existing variety graph that includes a plurality of nodes related to the existing varieties and links indicating relationships between the nodes, and is a trained graph of relationships between the nodes; and the request includes information indicating traits desired for the new variety; Breed improvement support method.

10. For computers, a process for receiving breeding requests; generating, based on the request, response information including information on existing varieties that are candidates for breeding to develop a new variety, using a trained model that has learned the relationship between the properties of existing varieties and the breeding process of the existing varieties; a process of generating evidence information including at least one of information on the breeding of the existing varieties to be crossbreeding candidates and information on the traits of the existing varieties to be crossbreeding candidates; a process of outputting the response information and the basis information; Execute The trained model is an existing variety graph that includes a plurality of nodes related to the existing varieties and links indicating relationships between the nodes, and is a trained graph of relationships between the nodes; and the request includes information indicating traits desired for the new variety; Breeding support program.

Images