A node screening-based substance recommendation method

Abstract

The present application relates to the field of data processing, and more particularly to a substance recommendation method based on node screening, comprising: screening out a first abnormal human body feature node and a first candidate node pathway based on first human body feature detection data, screening out a second abnormal human body feature node in combination with first connection relationship data, obtaining a first node distance between the second abnormal human body feature node and a substance node, screening out a second candidate node pathway based on an enrichment score of each node pathway, obtaining target feature data corresponding to the first human body feature node and screening out a target human body feature node according to second human body feature data and third human body feature data, and then obtaining a second node distance between the target human body feature node and the substance node and screening out a target substance node, wherein the influence of redundant node information on abnormal situation analysis is eliminated by screening the first human body feature node and the node pathway, and the accuracy of the substance recommendation result is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of data processing, in particular to a substance recommendation method based on node screening. BACKGROUND

[0002] Different users have different factors that change abnormally under abnormal conditions, resulting in complex and diverse abnormal data of users, making it difficult for users to filter and find effective solutions from a large number of abnormal solutions.

[0003] Traditional recommendation methods generally acquire a user graph under various abnormal conditions according to node characteristic data of multiple dimensions of the user and node characteristic data of the substance to be recommended, and analyze the node characteristic data in the user graph to recommend substances to the user. However, the abnormal data of the user is complex and diverse, and it is difficult to analyze the correlation and interaction between the node characteristics of multiple dimensions, resulting in low accuracy of the analysis of the correlation and interaction, and low accuracy of the recommendation of substances to the user, which is difficult to meet the requirements of the user.

[0004] Therefore, how to improve the accuracy of recommending substances to the target user has become a problem to be solved. SUMMARY

[0005] In view of the above technical problems, the technical scheme adopted by the present application is a substance recommendation method based on node screening, which comprises the following steps:

[0006] S100, a plurality of first human characteristic nodes in a target user graph, a second human characteristic node corresponding to each first human characteristic node, a third human characteristic node corresponding to each first human characteristic node, a plurality of substance nodes, a plurality of node pathways corresponding to all first human characteristic nodes, and first human characteristic detection data corresponding to all first human characteristic nodes, first connection relationship data corresponding to each first human characteristic node, second human characteristic data corresponding to each second human characteristic node, and third human characteristic data corresponding to each third human characteristic node are acquired.

[0007] S200, according to the first human characteristic detection data, a first abnormal human characteristic node is screened out from all the first human characteristic nodes, and the node pathway corresponding to the first abnormal human characteristic node is determined as a first candidate node pathway.

[0008] S300, according to the first connection relationship data, a second abnormal human characteristic node is screened out from the first abnormal human characteristic node corresponding to each first candidate node pathway.

[0009] S400, obtain the first node distance between each second abnormal human body feature node and each material node according to the positions of the second abnormal human body feature nodes and the material nodes in the target user graph.

[0010] S500, obtain the enrichment score of each node path, and screen the second candidate node path from all node paths according to the preset first score threshold and the preset second score threshold.

[0011] S600, obtain the target feature data corresponding to each first human body feature node according to the second human body feature data corresponding to each second human body feature node and the third human body feature data corresponding to each third human body feature node.

[0012] S700, screen the target human body feature node from the first human body feature nodes corresponding to each second candidate node path according to the target feature data, the preset first feature threshold and the preset second feature threshold.

[0013] S800, obtain the second node distance between each target human body feature node and each material node according to the positions of the target human body feature nodes and the material nodes in the target user graph.

[0014] S900, screen the target material node from all material nodes according to the first node distance, the second node distance and the preset distance threshold.

[0015] The present application has obvious beneficial effects compared with the prior art. The material recommendation method based on node screening provided by the present application can achieve considerable technical progress and practicality, and has wide industrial utilization value. At least the following beneficial effects are achieved: the first human body feature nodes are classified and screened according to the first human body feature detection data, the first abnormal human body feature nodes are screened from all the first human body feature nodes, and the node pathway corresponding to the first abnormal human body feature nodes is determined as the first candidate node pathway, thereby improving the pertinence and accuracy of material recommendation; the most downstream first abnormal human body feature node in the first candidate node pathway is regarded as a second abnormal human body feature node according to the first connection relationship data, so as to analyze the abnormal situation caused by the abnormal feature node of the target user and the effect of repairing the abnormal situation by the material node, thereby improving the accuracy of the analysis result; the first node distance between each second abnormal human body feature node and each material node is obtained according to the positions of the second abnormal human body feature nodes and the material nodes in the target user graph, so as to represent the effect of each material node on each second abnormal human body feature node; the enrichment score of each node pathway is obtained, so as to represent the enrichment degree of the first human body feature nodes corresponding to the node pathway at both ends of the sorting sequence of the preset reference node pathway set, and then represent the influence of the cooperative change of the first human body feature nodes in the corresponding node pathway on the preset phenotype state, the second candidate node pathway is screened from all the node pathways according to the preset first score threshold and the preset second score threshold, the target feature data corresponding to each first human body feature node is obtained according to the second human body feature data and the third human body feature data, the target human body feature nodes are screened from the first human body feature nodes corresponding to each second candidate node pathway by combining the preset first feature threshold and the preset second feature threshold, the influence of the first human body feature nodes with moderate expression on the abnormal situation analysis is eliminated, and the pertinence and accuracy of material recommendation are improved; the second node distance between each target human body feature node and each material node is obtained according to the positions of the target human body feature nodes and the material nodes in the target user graph, the target material node is screened from all the material nodes according to the first node distance, the second node distance and the preset distance threshold, and the accuracy of the material recommendation result is improved. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0017] Figure 1A flowchart of a substance recommendation method based on node screening provided for an embodiment of the present application is shown in the figure;

[0018] Figure 2 A flowchart of a substance recommendation method based on user graph provided for the second embodiment of the present application is shown in the figure;

[0019] Figure 3 A flowchart of an execution computer program of a substance recommendation system provided for the first embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0020] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.

[0021] It should be noted that the terms "first", "second" and the like in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or a chronological sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, method, substance or server including a series of steps or units does not have to be limited to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, substances or devices.

[0022] Embodiment one

[0023] The first embodiment of the present application provides a substance recommendation method based on node screening, as shown in the figure, the substance recommendation method based on node screening comprises the following steps: Figure 1

[0024] S100, obtaining a plurality of first human feature nodes in a target user graph, a second human feature node corresponding to each first human feature node, a third human feature node corresponding to each first human feature node, a plurality of substance nodes, a plurality of node pathways corresponding to all first human feature nodes, and first human feature detection data corresponding to all first human feature nodes, first connection relationship data corresponding to each first human feature node, second human feature data corresponding to each second human feature node, and third human feature data corresponding to each third human feature node.

[0025] ​The target user can refer to a user who needs to be recommended with relevant substances. The first human body feature node, the second human body feature node, and the third human body feature node can refer to nodes that change differently when the target user is in an abnormal situation, and can present different node abnormal situations.

[0026] The substance corresponding to the substance node refers to a recommendation target for recommending the target user. The substance corresponding to the substance node acts on the first human body feature node of the target user based on the second human body feature node, and then acts on the target user.

[0027] The first connection relationship data corresponding to the first human body feature node is used to represent the connection relationship and interaction relationship between the first human body feature nodes. For example, the first human body feature node can have an activating effect, an inhibiting effect, and the like on other first human body feature nodes. The node pathway is a pathway composed of the first human body feature nodes based on the first connection relationship data corresponding to the first human body feature nodes.

[0028] The first human body feature detection data corresponding to all the first human body feature nodes can refer to whole exome sequencing data of the target user, and is used to represent whether the first human body feature corresponding to the first human body feature node changes.

[0029] The second human body feature data corresponding to each second human body feature node can refer to the information expression amount corresponding to the second human body feature node.

[0030] The third human body feature data corresponding to each third human body feature node can refer to the information expression amount corresponding to the third human body feature node.

[0031] The target user graph represents the first human body feature node, the second human body feature node, the third human body feature node, the abnormal feature node, and the substance node corresponding to the target user, as well as the association relationship and interaction relationship between the nodes, and represents the abnormal situation of the target user from multiple dimensions, thereby providing a data basis for recommending substances to the target user.

[0032] In a specific embodiment, the target user graph is obtained by the following steps:

[0033] S1, obtaining the first human body feature node set A, the first human body feature data set α, the second human body feature data set β, the third human body feature data set γ, the abnormal feature node set E, the substance node set F corresponding to the target user, and the second human body feature node set B, the third human body feature node set C, and the first connection relationship data set G corresponding to the first human body feature node set A, and the second connection relationship data set H corresponding to the second human body feature node set B, the fourth connection relationship data set Y corresponding to the abnormal feature node set E, and the fifth connection relationship data set Z corresponding to the substance node set F, wherein A = {A1, A2, …, Ai , ..., A I}, A i It refers to the i-th first human feature node, B = {B1, B2, ..., B...} i , ..., B I}, B i Refers to A i The corresponding second human feature node, C = {C1, C2, ..., C...} i , ..., C I}, C i Refers to A i The corresponding third human feature node, G is used to represent the connection relationship between the first human feature nodes, H is used to represent the connection relationship between the second human feature nodes, Y is used to represent the connection relationship between the abnormal feature node and the first human feature node, the second human feature node and the third human feature node, Z is used to represent the connection relationship between the abnormal feature node and the second human feature node, i = 1, 2, ..., I, u(i) = 1, 2, ..., i-1, i+1, ..., I, where I refers to the total number of first human feature nodes, j = 1, 2, ..., n, where n refers to the total number of change categories.

[0034] S2, based on the first human feature node set A, the second human feature node set B, and the third human feature node set C, obtain the first user graph set P = {P1, P2, ..., P...} i ..., P I}, where the i-th first user graph P i By analyzing A i B i and C i The two are connected in pairs.

[0035] Among them, the first human feature node A i There exists a corresponding second human feature node B. i And a corresponding third human feature node C i First, regarding A i B i and C i By connecting each pair of nodes, the corresponding first user graph P can be obtained. i For each of the I first human feature nodes, a first user graph set P = {P1, P2, ..., P} can be obtained. i ..., P I This serves as the basis for analyzing the node relationships of the target user.

[0036] The first human body feature node of the target user, the corresponding second human body feature node and the corresponding third human body feature node are connected in pairs to obtain a corresponding first user graph, which represents the association relationship between the first human body feature node, the corresponding second human body feature node and the corresponding third human body feature node, and provides a data basis for material recommendation for the target user.

[0037] S3, according to the first connection relationship data set G and the second connection relationship data set H, connecting all the first user graphs in the first user graph set P to obtain a second user graph.

[0038] The first human body feature node is connected according to the first connection relationship data between the first human body feature nodes, the second human body feature node is connected according to the second connection relationship data between the second human body feature nodes, and the first user graph is combined, which can complete the connection between all the first human body feature nodes, the second human body feature nodes and the third human body features, and obtain the second user graph, improve the association degree between the first human body feature nodes, the second human body feature nodes and the third human body features, and improve the representation accuracy of the association relationship and the interaction relationship between the nodes.

[0039] In a specific embodiment, the first connection relationship data set G={G1, G2, …, G i , …, G I}, G i ={G i1 , G i2 , …, G iu(i) , …, G iI}, G iu(i) represents the first connection relationship data between A i and A u(i) , and the first connection relationship data is used to represent whether there is a first connection line segment, the arrow direction of the first connection line segment and the first preset category of the first connection line segment, S3 further includes the following steps:

[0040] S311, obtaining the first connection relationship data G i between A u(i) and A iu(i) ={GX iu(i) 1 , GX iu(i) 2}, wherein GX iu(i) 1 =GX iu(i) 2 =0 represents that there is no first connection line segment between A i and A u(i) , GX iu(i)1 =1 means that there is a first connection line segment between A i and A u(i) and the arrow direction is from A i to A u(i) , GX iu(i) 1 =2 means that there is a first connection line segment between A i and A u(i) and the arrow direction is from A u(i) to A i , GX iu(i) 2 =ρ means that there is a first connection line segment between A i and A u(i) and the first connection line segment belongs to the ρth first preset category, ρ=1, 2, …, μ, μ is the total number of the first preset categories of the first connection line segment.

[0041] S312, if GX iu(i) 1 =GX iu(i) 2 =0, A i and A u(i) are not connected.

[0042] S313, if GX iu(i) 1 =1 and GX iu(i) 2 =ρ, A i and A u(i) are connected according to the first connection line segment whose arrow direction is from A i to A u(i) and which belongs to the ρth first preset category.

[0043] S314, if GX iu(i) 1 =2 and GX iu(i) 2 =ρ, A u(i) and A i are connected according to the first connection line segment whose arrow direction is from A i to A u(i) and which belongs to the ρth first preset category.

[0044] wherein the first preset category can be set by the implementer according to actual conditions, for example, the first preset category can be a binding, activation, inhibition, or other interactive relationship category.

[0045] The above-mentioned method determines whether there is a first connecting line segment and the arrow direction and category of the first connecting line segment between the first human feature nodes based on the first connection relationship data between the first human feature nodes, thereby completing the connection between the first human feature nodes, improving the correlation between the first human feature nodes in the second user graph, and thus improving the accuracy of the representation of the correlation and interaction relationship between nodes.

[0046] In one specific implementation, the second connection relation dataset H = {H1, H2, ..., H...} i H I}, H i ={H i1 H i2 H iu(i) H iI}, H iu(i) It refers to B i and B u(i) The second connection relationship data between them, which is used to indicate whether a second connection segment exists and the confidence level of the second connection segment, S3 also includes the following steps:

[0047] S321, obtain B i and B u(i) The second connection relationship data H between them iu(i) , where H iu(i) =0 indicates B i and B u(i) There is no second connecting line segment between them, H iu(i) >0 indicates B i and B u(i) There exists a second connecting segment and the confidence level corresponding to the second connecting segment is equal to H. iu(i) .

[0048] S322, if H iu(i) =0, then B is not connected. i and B u(i) .

[0049] S323, if H iu(i) If the value is greater than 0, then the corresponding confidence level is equal to H. iu(i) The second connecting line segment connects to B i and B u(i) .

[0050] The confidence level is used to characterize the degree of credibility of the existence of a second connecting line segment between two corresponding second human feature nodes, and can be obtained by the implementer from a preset database.

[0051] According to the second connection relationship data between the second human body feature nodes, it is determined whether the second connection line segments exist between the second human body feature nodes and the confidence degrees corresponding to the second connection line segments, the connection between the second human body feature nodes is completed, the correlation degree between the second human body feature nodes in the second user graph is improved, and therefore the representation accuracy of the correlation and interaction between the nodes is improved.

[0052] S4, according to the first human body feature data set α, the second human body feature data set β and the third human body feature data set γ, updating the node feature data of all the first human body feature nodes, the second human body feature nodes and the third human body feature nodes in the second user graph, obtaining the third user graph corresponding to the target user.

[0053] The specific change category can be set by the implementer according to the actual situation. For example, in the embodiment, the traversal category can be a change category such as frame deletion mutation, frame insertion mutation, frame deletion, frame insertion, missense mutation, nonsense mutation, stop codon mutation, splice site mutation, start codon mutation, base insertion, etc.

[0054] The node feature data corresponding to the second human body feature node can be the information expression amount corresponding to the second human body feature node. The node feature data corresponding to the third human body feature node can be the information expression amount corresponding to the third human body feature node.

[0055] The initial node feature data of all the first human body feature nodes, the second human body feature nodes and the third human body feature nodes in the second user graph is 0.

[0056] After obtaining the first human body feature data set α, the second human body feature data set β and the third human body feature data set γ corresponding to the target user, the node feature data of all the first human body feature nodes, the second human body feature nodes and the third human body feature nodes in the second user graph is updated, the third user graph corresponding to the target user is obtained, the representation degree of the human body feature data of the target user by the third user graph is improved, and when the third user graph is used as a data basis for material recommendation for the target user, the accuracy of the material recommendation result is improved.

[0057] S5, according to the fourth connection relationship data set Y and the fifth connection relationship data set Z, adding all the abnormal feature nodes in the abnormal feature node set E and all the material nodes in the material node set F to the third user graph, obtaining the target user graph.

[0058] In a specific embodiment, S5 further includes the following steps:

[0059] S531, obtain the substance node set F and the fifth connection relationship data set Z corresponding to F, wherein F={F1, F2, …, F t , …, F T}, F t refers to the t-th substance node, Z={Z1, Z2, …, Z t , …, Z T}, F T refers to the fifth connection relationship data list Z t corresponding to F t . t 1 t 2}, Z t 1 ε refers to the ε-th second human body feature node connected by F t , refers to the η-th abnormal feature node connected by E t , t=1, 2, …, T, T refers to the total number of substance nodes, ε=1, 2, …, ζ(t), ζ(t) refers to the total number of second human body feature nodes connected by F t , η=1, 2, …, λ(t), λ(t) refers to the total number of abnormal feature nodes connected by F t .

[0060] S532, add all substance nodes in F to the second user sub-graph according to Z, and obtain the target user graph.

[0061] , wherein the substance corresponding to the substance node refers to a recommendation target for recommending the target user, and the substance corresponding to the substance node acts on the abnormal feature node of the target user based on the second human body feature node, and then acts on the target user.

[0062] Therefore, the embodiment can connect the substance node, the corresponding second human body feature node and the abnormal feature node according to the fifth connection relationship corresponding to the substance node, to represent the association relationship and the interaction relationship between the substance node and the second human body feature node and the abnormal feature node, thereby improving the representation accuracy of the association relationship and the interaction relationship between the nodes in the target user graph.

[0063] In a specific embodiment, the node path is obtained by the following steps:

[0064] S101, for the i-th first human body feature node A i , if G i1 1 =G i2 1 =…=Giu(i) 1 ≠…≠G iI 1 ≠1, A i is updated to the first level target node.

[0065] S102, according to G i , the first human feature node pointing to A i is updated to the corresponding second level target node A i .

[0066] S103, according to G, the first human feature node pointing to A i corresponding to the ω level target node is updated to A i corresponding to the ω+1 level target node, until the total number of A i corresponding to the ω+1 level target node is equal to 0, wherein ω is an integer greater than 1.

[0067] S104, in the order of A i , A i corresponding to the second level target node, A i corresponding to the third level target node to A i corresponding to the ω level target node, A i corresponding to the node path is obtained.

[0068] S105, traversing the first human feature node set A, all node paths are obtained.

[0069] When there is a first connection line segment between two first human feature nodes in the target user graph, it can represent that there is an interaction relationship between the corresponding two first human feature nodes, and the arrow direction of the first connection line segment can represent that the first human feature node has binding, activation, inhibition and other interactions with the pointed first human feature node.

[0070] At the same time, the arrow direction of the first connection line segment can represent the direction from the upstream node to the downstream node in the corresponding node path, so as to trace the upstream node from the most downstream first level target node to obtain the corresponding second level target node, third level target node and other upstream nodes of each level, and then the candidate node path can be obtained to further serve as the data basis of material recommendation.

[0071] Specifically, the first human feature node pointing to the second level target node is determined as the third level target node, the first human feature node pointing to the third level target node is determined as the fourth level target node, and so on, the first human feature node pointing to the ω level target node is determined as the ω+1 level target node, until the total number of the ω+1 level target nodes is equal to 0, which indicates that all the upstream nodes corresponding to the first level target nodes have been determined, and then the acquisition of the target nodes is stopped.

[0072] Therefore, in the order of the first level target nodes, the second level target nodes, the third level target nodes, and so on, the corresponding candidate node paths can be acquired in reverse. Further, by traversing all the first target nodes, all the node paths in the target user graph can be acquired as the basis for the material recommendation for the target user, and the influence of the redundant node information in the target user graph on the material recommendation result is eliminated, thereby improving the accuracy of the material recommendation result.

[0073] S200, according to the first human feature detection data, screening the first abnormal human feature nodes from all the first human feature nodes, and determining the node path corresponding to the first abnormal human feature nodes as the first candidate node path.

[0074] In a specific embodiment, S200 further includes the following steps:

[0075] S201, according to the first human feature detection data, judging the state detection category corresponding to each first human feature node, wherein the state detection category is abnormal or not abnormal.

[0076] S202, when the state detection category corresponding to the first human feature node is abnormal, determining the corresponding first human feature node as the first abnormal human feature node.

[0077] The first human feature detection data can represent whether the first human feature corresponding to the first human feature node changes, and therefore, according to the first human feature detection data, the state detection category corresponding to each first human feature node can be judged as abnormal or not abnormal, and when the state detection category corresponding to the first human feature node is abnormal, the corresponding first human feature node can be determined as the first abnormal human feature node.

[0078] The above, according to the first human feature detection data, the first human feature nodes are classified and screened, and the abnormal first human feature nodes are selected as the analysis objects, and the influence of the non-abnormal node paths of the non-abnormal first human feature nodes on the abnormal situation analysis is eliminated, thereby the abnormal situation analysis of the target user can be more efficiently and accurately performed, and the pertinence and accuracy of the material recommendation are improved.

[0079] S300, according to the first connection relationship data, filtering out the second abnormal human body feature node from the first abnormal human body feature node corresponding to each first candidate node path.

[0080] In a specific embodiment, S300 further comprises the following steps:

[0081] S301, for any first candidate node path, obtaining the level sequence number corresponding to each first abnormal human body feature node in the first candidate node path.

[0082] S302, determining the first abnormal human body feature node corresponding to the smallest level sequence number in the first candidate node path as the second abnormal human body feature node corresponding to the first candidate node path.

[0083] In each candidate node path, the upstream first human body feature node will bind, activate, inhibit and other interactive behaviors to the downstream first human body feature node, so that the upstream first human body feature node will also affect the abnormal situation of the first human body feature node when the abnormal feature node causes the abnormal situation of the downstream first human body feature node and the material node repairs the abnormal situation of the downstream first human body feature node. Therefore, when analyzing the abnormal situation caused by the abnormal feature node of the target user and analyzing the effect of repairing the abnormal situation by the material node, the most downstream first abnormal human body feature node in the first candidate node path is regarded as the most concerned node.

[0084] The level sequence number corresponding to each first abnormal human body feature node can represent the upstream and downstream situation of the position of the first abnormal human body feature node in the corresponding first candidate node path, and the smaller the level sequence number is, the more downstream the position of the first abnormal human body feature node in the corresponding first candidate node path is. Then the first abnormal human body feature node corresponding to the smallest level sequence number in the first candidate node path is determined as the second abnormal human body feature node corresponding to the first candidate node path.

[0085] Based on the interactive behavior of the upstream first human body feature node to the downstream first human body feature node, the most downstream first abnormal human body feature node in the first candidate node path is regarded as the most concerned second abnormal human body feature node to analyze the abnormal situation caused by the abnormal feature node of the target user and analyze the effect of repairing the abnormal situation by the material node, which improves the accuracy of the analysis result.

[0086] S400, according to the position of the second abnormal human body feature node and the material node in the target user graph, obtaining the first node distance between each second abnormal human body feature node and each material node.

[0087] In a specific embodiment, S400 further comprises the following steps:

[0088] S401, obtaining a t-th material node F t a connected ε-th second human feature node Z t 1 ε and Z t 1 ε a connected first human feature node, Z t 1 ε a connected first human feature node F t a corresponding first reference human feature node.

[0089] S402, for any second abnormal feature node, if the second abnormal feature node and F t the corresponding first reference human feature node are on the same node path, obtaining a first level sequence number of the second abnormal feature node on the node path and F t a second level sequence number of the corresponding first reference human feature node on the node path.

[0090] S403, according to the first level sequence number and the second level sequence number, obtaining a first node distance between the second abnormal feature node and F t .

[0091] S404, if the second abnormal feature node and F t the corresponding first reference human feature node are not on the same node path, determining a preset node distance as the first node distance between the second abnormal feature node and F t .

[0092] Wherein, the material corresponding to the material node refers to the recommendation target for recommending the target user, and the material corresponding to the material node acts on the first human feature node of the target user, and then acts on the target user based on the second human feature node.

[0093] Therefore, the node distance between the second abnormal human feature node and the material node can be converted into the node distance between the second human feature node connected to the second abnormal human feature node and the first human feature node connected to the second human feature node.

[0094] When the second abnormal feature node and the corresponding first reference human feature node are on the same node path, the corresponding first node distance can be represented according to the upstream and downstream position difference of the node in the node path.

[0095] When the second abnormal feature node and the corresponding first reference human feature node are not on the same node path, it indicates that the corresponding material node has a smaller effect on the second abnormal feature node, that is, the first node distance between the second abnormal feature node and the corresponding material node is farther, and at this time, the preset node distance is determined as the corresponding first node distance.

[0096] The preset node distance can be set according to actual conditions, for example, the preset node distance can be the largest level number.

[0097] According to the positions of the second abnormal human feature nodes and the material nodes in the target user graph, the first node distance between each second abnormal human feature node and each material node is obtained to represent the effect of each material node on each second abnormal human feature node, and serves as the basis for material recommendation, thereby improving the accuracy of the material recommendation result.

[0098] S500, obtain the enrichment score of each node path, and filter out the second candidate node path from all node paths according to the preset first score threshold and the preset second score threshold.

[0099] The phenotype refers to the traits and characteristics (morphology, structure, physiology, biochemistry, behavior, etc.) exhibited by an individual under certain environmental conditions. The preset phenotype state can be set by the implementer according to actual needs.

[0100] The embodiment can obtain the reference node path corresponding to each type of phenotype state, and then filter out the reference node path corresponding to the preset phenotype state and form the preset reference node path set. Then, based on the node path and the reference node path set, the enrichment score of each node path is obtained to represent the enrichment degree of the two ends of the sorting sequence of the preset reference node path set corresponding to the first human feature node in the node path, and then to represent the influence of the coordinated change of the first human feature node in the corresponding node path on the preset phenotype state, so that the candidate node path with a larger or smaller enrichment score can be selected as the second candidate node path with a higher attention degree for material recommendation, so that the recommended material has a higher matching degree with the preset phenotype state and the actual needs of the implementer, thereby improving the accuracy of the recommendation result.

[0101] Specifically, the expression degree of each first human feature node in the preset reference node path set under the preset phenotype state is obtained, and the first human feature nodes in the reference node path set are sorted in descending order of the expression degree, and the sorting sequence of the preset reference node path set can be obtained.

[0102] The calculation method of the enrichment score is a known technical means in the art, which is not repeated here.

[0103] The enrichment score is obtained to represent the enrichment degree of the first human characteristic node corresponding to the node pathway at both ends of the ranking sequence of the preset reference node pathway set, thereby representing the influence of the collaborative change of the first human characteristic node in the corresponding node pathway on the preset phenotype state, and the second candidate node pathway with higher attention and higher matching degree to the actual needs of the implementer is screened from the node pathway, thereby improving the accuracy of the material recommendation result.

[0104] In a specific embodiment, the preset second score threshold is smaller than the preset first score threshold, and S500 further includes the following steps:

[0105] For any node pathway, if the enrichment score corresponding to the node pathway is greater than the preset first score threshold or the enrichment score corresponding to the node pathway is less than the preset second score threshold, the node pathway is determined as the second candidate node pathway.

[0106] S600, according to the second human characteristic data corresponding to each second human characteristic node and the third human characteristic data corresponding to each third human characteristic node, the target characteristic data corresponding to each first human characteristic node is obtained.

[0107] Each first human characteristic node has a corresponding second human characteristic node and a corresponding second human characteristic node, which can form a corresponding human characteristic triple.

[0108] The second human characteristic data can refer to the information expression amount corresponding to the second human characteristic node, and the third human characteristic data can refer to the information expression amount corresponding to the third human characteristic node. For each human characteristic triple, according to the second human characteristic data corresponding to the second human characteristic node and the third human characteristic data corresponding to the third human characteristic node, the target characteristic data corresponding to the first human characteristic node can be obtained to represent the expression amount of the first human characteristic node.

[0109] For example, the target characteristic data can be the product of the first preset priority and the second human characteristic data, plus the product of the second preset priority and the third human characteristic data. The first preset priority and the second preset priority are respectively used to represent the importance of the second human characteristic node and the third human characteristic node in expressing the node data of the corresponding first human characteristic node.

[0110] S700, according to the target characteristic data, the preset first feature threshold and the preset second feature threshold, the target human characteristic node is screened from the first human characteristic node corresponding to each second candidate node pathway.

[0111] In one specific embodiment, where the preset second feature threshold is less than the preset first feature threshold, S700 further includes the following steps:

[0112] For any target feature data, if the target feature data is greater than a preset first feature threshold, or if the target feature data is less than a preset second feature threshold, then the first human feature node corresponding to the target feature data is determined as the target human feature node.

[0113] In analyzing anomalies in target users, selecting first human feature nodes with high or low expression levels as high-attention target human feature nodes, and eliminating the influence of first human feature nodes with medium expression levels on anomaly analysis, can more efficiently and accurately analyze anomalies in target users, thus improving the targeting and accuracy of substance recommendations.

[0114] S800 obtains the second node distance between each target human feature node and each material node based on the position of the target human feature node and material node in the target user map.

[0115] In one specific embodiment, S800 further includes the following steps:

[0116] S801, for any target human feature node, if the target human feature node and F t If the corresponding first reference human feature node is on the same node path, then the third-level sequence number of the target human feature node on the node path is obtained.

[0117] S802, based on the third-level sequence number and the second-level sequence number, obtain the target human body feature nodes and F. t The distance between the second node.

[0118] S803, if the target human body feature nodes and F t If the corresponding first reference human feature nodes are not on the same node path, then the preset node distance is determined as the distance between the target human feature node and F. t The distance between the second node.

[0119] Specifically, the node distance between the target human feature node and the material node can be transformed into the node distance between the target human feature node and the first human feature node connected to the second human feature node connected to the material node.

[0120] When the target human feature node and the corresponding first reference human feature node are on the same node path, the distance to the corresponding second node can be characterized by the difference in the upstream and downstream positions of the node in the node path.

[0121] When the target human feature node and the corresponding first reference human feature node are not on the same node path, it indicates that the corresponding material node has a smaller effect on the target human feature node, that is, the second node distance between the target human feature node and the corresponding material node is farther, and the preset node distance is determined as the corresponding second node distance.

[0122] According to the positions of the target human feature nodes and the material nodes in the target user graph, the second node distance between each target human feature node and each material node is obtained to represent the effect of each material node on each target human feature node and serve as the basis for material recommendation, thereby improving the accuracy of the material recommendation result.

[0123] S900, filtering out the target material node from all the material nodes according to the first node distance, the second node distance, and the preset distance threshold.

[0124] In a specific embodiment, S900 further includes the following steps:

[0125] S901, filtering out the first recommended material node from all the material nodes according to the first node distance and the preset first distance threshold, and obtaining the first recommended material node set.

[0126] S902, filtering out the second recommended material node from all the material nodes according to the second node distance and the preset first distance threshold, and obtaining the second recommended material node set.

[0127] S903, determining the intersection of the first recommended material node set and the second recommended material node set as the set corresponding to the target material node.

[0128] When the first node distance is smaller than the preset first distance threshold, the material node corresponding to the first node distance is determined as the first recommended material node.

[0129] When the second node distance is smaller than the preset first distance threshold, the material node corresponding to the second node distance is determined as the second recommended material node.

[0130] When the material node is neither the first recommended material node nor the second recommended material node, the material node is determined as the target material node, and the material corresponding to the target material node is recommended to the target user.

[0131] The specific value of the first distance threshold can be set by the implementer according to the actual situation.

[0132] According to the first human body feature detection data, the first human body feature nodes are classified and screened, the first abnormal human body feature nodes are screened from all the first human body feature nodes, and the node path corresponding to the first abnormal human body feature nodes is determined as the first candidate node path, thereby improving the pertinence and accuracy of the substance recommendation; according to the first connection relationship data, the most downstream first abnormal human body feature node in the first candidate node path is regarded as a second abnormal human body feature node, so as to analyze the abnormal situation caused by the abnormal feature node of the target user and analyze the effect of repairing the abnormal situation by the substance node, thereby improving the accuracy of the analysis result; according to the positions of the second abnormal human body feature nodes and the substance nodes in the target user graph, the first node distance between each second abnormal human body feature node and each substance node is obtained to represent the effect of each substance node on each second abnormal human body feature node; the enrichment score of each node path is obtained to represent the enrichment degree of the first human body feature nodes corresponding to the node path at both ends of the sorting sequence of the preset reference node path set, thereby representing the influence of the cooperative change of the first human body feature nodes in the corresponding node path on the preset phenotype state, according to the preset first score threshold and the preset second score threshold, the second candidate node path is screened from all the node paths, according to the second human body feature data and the third human body feature data, the target feature data corresponding to each first human body feature node is obtained, the target human body feature nodes are screened from the first human body feature nodes corresponding to each second candidate node path by combining the preset first feature threshold and the preset second feature threshold, the influence of the first human body feature nodes with moderate expression on the abnormal situation analysis is eliminated, and the pertinence and accuracy of the substance recommendation are improved; according to the positions of the target human body feature nodes and the substance nodes in the target user graph, the second node distance between each target human body feature node and each substance node is obtained, the target substance node is screened from all the substance nodes according to the first node distance, the second node distance and the preset distance threshold, and the accuracy of the substance recommendation result is improved.

[0133] Embodiment Two

[0134] On the basis of the above-mentioned embodiment one, the embodiment two provides a substance recommendation method based on a user graph, as shown in the following formula (I): Figure 2 The substance recommendation method based on the user graph includes:

[0135] S1, obtain the first human feature node set A, the first human feature dataset α, the second human feature dataset β, the third human feature dataset γ, the abnormal feature node set E, and the material node set F corresponding to the target user; the second human feature node set B, the third human feature node set C, and the first connection relationship dataset G corresponding to the first human feature node set A; and the second connection relationship dataset H corresponding to the second human feature node set B, the fourth connection relationship dataset Y corresponding to the abnormal feature node set E, and the fifth connection relationship dataset Z corresponding to the material node set F, where A = {A1, A2, ..., A...} i , ..., A I}, A i It refers to the i-th first human feature node, B = {B1, B2, ..., B...} i , ..., B I}, B i Refers to A i The corresponding second human feature node, C = {C1, C2, ..., C...} i , ..., C I}, C i Refers to A i The corresponding third human feature node, G is used to represent the connection relationship between the first human feature nodes, H is used to represent the connection relationship between the second human feature nodes, Y is used to represent the connection relationship between the abnormal feature node and the first human feature node, the second human feature node and the third human feature node, Z is used to represent the connection relationship between the abnormal feature node and the second human feature node, i = 1, 2, ..., I, u(i) = 1, 2, ..., i-1, i+1, ..., I, where I refers to the total number of first human feature nodes, j = 1, 2, ..., n, where n refers to the total number of change categories.

[0136] S2, based on the first human feature node set A, the second human feature node set B, and the third human feature node set C, obtain the first user graph set P = {P1, P2, ..., P...} i ..., P I}, where the i-th first user graph P i By analyzing A i B i and C i The two are connected in pairs.

[0137] S3. Based on the first connection relationship dataset G and the second connection relationship dataset H, connect all the first user graphs in the first user graph set P to obtain the second user graph.

[0138] In one specific implementation, the first connection relation dataset G = {G1, G2, ..., G...} i , ..., GI}, G i ={G i1 G i2 , ..., G iu(i) , ..., G iI}, G iu(i) Refers to A i and A u(i) The first connection relationship data between them, which is used to indicate whether there is a first connecting line segment, the direction of the arrow of the first connecting line segment, and the first preset category of the first connecting line segment, S3 also includes the following steps:

[0139] S311, obtain A i and A u(i) The first connection relationship data G between them iu(i) ={GX iu(i) 1 GX iu(i) 2}, where GX iu(i) 1 =GX iu(i) 2 =0 means A i and A u(i) There is no first connecting line segment between them, GX iu(i) 1 =1 means A i and A u(i) There is a first connecting line segment, and the arrow points from A. i Pointing to A u(i) GX iu(i) 1 =2 means A i and A u(i) There is a first connecting line segment, and the arrow points from A. u(i) Pointing to A i GX iu(i) 2 =ρ represents A i and A u(i) There exists a first connecting line segment and the category of the first connecting line segment is the ρth first preset category, where ρ = 1, 2, ..., μ, and μ refers to the total number of the first preset categories of the first connecting line segment.

[0140] S312, if GX iu(i) 1 =GX iu(i) 2 =0, then A is not connected. i and A u(i) .

[0141] S313, if GX iu(i) 1 =1 and GXiu(i) 2 =ρ, then according to the arrow direction, it is from A i Pointing to A u(i) And the first connecting line segment A belongs to the first preset category of type ρ. i and A u(i) .

[0142] S314, if GX iu(i) 1 =2 and GX iu(i) 2 =ρ, then according to the arrow direction, it is from A u(i) Pointing to A i And the first connecting line segment A belongs to the first preset category of type ρ. i and A u(i) .

[0143] In one specific implementation, the second connection relation dataset H = {H1, H2, ..., H...} i H I}, H i ={H i1 H i2 H iu(i) H iI}, H iu(i) It refers to B i and B u(i) The second connection relationship data between them, which is used to indicate whether a second connection segment exists and the confidence level of the second connection segment, S3 also includes the following steps:

[0144] S321, obtain B i and B u(i) The second connection relationship data H between them iu(i) H iu(i) =0 indicates B i and B u(i) There is no second connecting line segment between them, H iu(i) >0 indicates B i and B u(i) There exists a second connecting segment and the confidence level corresponding to the second connecting segment is equal to H. iu(i) .

[0145] S322, if H iu(i) =0, then B is not connected. i and B u(i) .

[0146] S323, if H iu(i) If the value is greater than 0, then the corresponding confidence level is equal to H. iu(i) The second connecting line segment connects to B i and Bu(i) .

[0147] Wherein, the confidence is used to represent the confidence degree of the existence of the second connection line segment between the corresponding two second human body feature nodes, which can be obtained by the implementer from the preset database.

[0148] The above, according to the second connection relationship data between the second human body feature nodes, judges whether the second connection line segment exists between the second human body feature nodes and the confidence corresponding to the second connection line segment, completes the connection between the second human body feature nodes, improves the correlation degree between the second human body feature nodes in the second user graph, and thus improves the representation accuracy of the correlation and interaction between the nodes.

[0149] S4, according to the first human body feature data set α, the second human body feature data set β and the third human body feature data set γ, updates the node feature data of all first human body feature nodes, second human body feature nodes and third human body feature nodes in the second user graph, and obtains the third user graph corresponding to the target user.

[0150] In a specific embodiment, α={α1, α2, …, α i , …, α m}, α i ={α i1 , α i2 , …, α ij , …, α in}, α ij is the number of changes of the jth change category corresponding to the ith first human body feature node, β={β1, β2, …, β i , …, β m}, β i is the node feature data corresponding to the ith second human body feature node, γ={γ1, γ2, …, γ i , …, γ m}, γ i is the node feature data corresponding to the ith third human body feature node, and S4 further includes the following steps:

[0151] S411, the node feature data corresponding to the ith first human body feature node in the second user graph is updated to α i .

[0152] S412, the node feature data corresponding to the ith second human body feature node in the second user graph is updated to β i .

[0153] S413, the node feature data corresponding to the ith third human body feature node in the second user graph is updated to γ i .

[0154] S5. Based on the fourth connection relationship dataset Y and the fifth connection relationship dataset Z, add all abnormal feature nodes in the abnormal feature node set E and all material nodes in the material node set F to the third user graph to obtain the target user graph.

[0155] In one specific embodiment, S5 further includes the following steps:

[0156] S521, Obtain the abnormal feature node set E and the corresponding fourth connection relation dataset Y, where E = {E1, E2, ..., E...} k , ..., E K}, E k It refers to the k-th anomalous feature node, Y = {Y1, Y2, ..., Y}. k , ..., Y K}, E k The corresponding fourth connection relationship data list Y k ={Y k 1 Y k 2 Y k 3}, It refers to E k The q-th first human feature node is connected. It refers to E k The r-th second human feature node is connected. It refers to E k The connection is to the s-th third human feature node, k = 1, 2, ..., K, where K is the total number of abnormal feature nodes, and q = 1, 2, ..., Q(k), where Q(k) is the E k The total number of connected first human feature nodes, r = 1, 2, ..., R(k), where R(k) refers to E k The total number of connected second human feature nodes, s = 1, 2, ..., S(k), where S(k) refers to E k The total number of connected third-party human feature nodes.

[0157] S522, add all abnormal feature nodes in E to the first user subgraph according to Y, and obtain the second user subgraph.

[0158] Among them, abnormal feature nodes can refer to feature nodes corresponding to abnormal situations. The abnormal situations corresponding to abnormal feature nodes will affect the first human feature node, the second human feature node, and the third human feature node.

[0159] Therefore, the embodiment can connect the abnormal feature node, all the corresponding first human body feature nodes, the second human body feature nodes and the third human body feature nodes according to the fourth connection relationship corresponding to the abnormal feature node, so as to represent the association relationship and the interaction relationship between the abnormal feature node and the first human body feature nodes, the second human body feature nodes and the third human body feature nodes, thereby improving the representation accuracy of the association relationship and the interaction relationship between the nodes in the target user graph.

[0160] In a specific embodiment, S5 further comprises the following steps:

[0161] S531, obtaining the substance node set F and the fifth connection relationship data set Z corresponding to F, wherein F={F1, F2, …, F t , …, F T}, F t refers to the tthsubstance node, Z={Z1, Z2, …, Z t , …, Z T}, F t corresponding to the fifth connection relationship data table Z t ={Z t 1 , Z t 2}, refers to the εthsecond human body feature node connected by F t , refers to the ηthabnormal feature node connected by E t , t=1, 2, …, T, T refers to the total number of substance nodes, ε=1, 2, …, ζ(t), ζ(t) refers to the total number of second human body feature nodes connected by F t ,

[0162] η=1, 2, …, λ(t), λ(t) refers to the total number of abnormal feature nodes connected by F t .

[0163] S6, inputting the target user graph into a preset recommendation model to obtain a target substance recommendation result corresponding to the target user.

[0164] The preset recommendation model is used to analyze the nodes in the target user graph, the connection relationship between the nodes, and the node feature data of the nodes, and output a predicted connection value between the abnormal feature nodes and the material nodes, to represent the degree of association and interaction between the abnormal feature nodes and the material nodes, and then filter out the material nodes with a predicted connection value greater than a preset connection value threshold, determine the material corresponding to the filtered material nodes as a target material corresponding to the target user, and thus obtain a target material recommendation result, thereby improving the accuracy of the target material recommendation result.

[0165] The preset recommendation model can be obtained and trained by the implementer according to the actual situation. For example, the preset recommendation model can include a preset first encoding model, a preset second encoding model, and a preset prediction model.

[0166] According to the first human feature node set A corresponding to the target user, the second human feature node set B corresponding to A, and the third human feature node set C corresponding to A, the first user graph set P is obtained, the first user graph set P is connected according to the first connection relationship data set G corresponding to A and the second connection relationship data set H corresponding to B, and the second user graph is obtained, which improves the association degree between the first human feature node, the second human feature node, and the third human feature node in the second user graph; the node feature data of all the first human feature nodes, the second human feature nodes, and the third human feature nodes in the second user graph is updated according to the first human feature data set α corresponding to the target user, the second human feature data set β, and the third human feature data set γ, and the third user graph corresponding to the target user is obtained, which improves the representation degree of the human feature data of the target user by the third user graph; all the abnormal feature nodes in E and all the material nodes in F are added to the third user graph according to the fourth connection relationship data set Y corresponding to the abnormal feature node set E and the fifth connection relationship data set Z corresponding to the material node set F, and the target user graph is obtained, which displays the first human feature node, the second human feature node, the third human feature node, the abnormal feature node, and the material node, and the association relationship and the interaction relationship between the nodes in the target user graph, and improves the representation accuracy of the association relationship and the interaction relationship between the nodes in the target user graph; the target user graph is input into the preset recommendation model, and the target material recommendation result corresponding to the target user is obtained, which improves the accuracy of the target material recommendation result.

[0167] Embodiment Three

[0168] On the basis of the above embodiment one and embodiment two, the embodiment three provides a substance recommendation system, the substance recommendation system comprises a processor and a memory storing a computer program, the memory further stores a preset third user graph corresponding to a target user, a preset target user graph, a preset first encoding model, a preset second encoding model and a preset prediction model, when the computer program is executed by the processor, the following steps are realized, such as Figure 3 As shown in the following formula:

[0169] S10, the first node connection matrix ψ0, the first human body feature data matrix ψ1, the second human body feature data matrix ψ2 and the third human body feature data matrix ψ3 corresponding to the target user are obtained according to the preset third user graph, wherein ψ0 includes the first connection relationship data between all first human body feature nodes corresponding to the target user, ψ1 includes the node feature data of all first human body feature nodes corresponding to the target user, ψ2 includes the node feature data of all second human body feature nodes corresponding to the target user, and ψ3 includes the node feature data of all third human body feature nodes corresponding to the target user.

[0170] Wherein, the first node connection matrix ψ0 is used to represent the connection relationship between any two first human body feature nodes, the first human body feature data matrix ψ1 is used to represent the change situation corresponding to each first human body feature node, including the change category situation and the change degree situation, the second human body feature data matrix ψ2 is used to represent the expression amount situation corresponding to each second human body feature node, and the third human body feature data matrix ψ3 is used to represent the expression amount situation corresponding to each third human body feature node, then the abnormal situation of the target user can be represented in multiple dimensions according to the third user graph and the corresponding ψ0, ψ1, ψ2 and ψ3, so as to provide data basis for the target user to carry out substance recommendation.

[0171] In a specific embodiment, the first node connection matrix ψ0 is obtained by the following steps:

[0172] S011, the first connection relationship data G between A and A is obtained i u(i) iu(i) = {GX iu(i) 1 , GX iu(i) 2}, wherein GX iu(i) 1 = GX iu(i) 2 = 0 represents that there is no first connection line segment between A and A i u(i) , GX iu(i) 1 = 1 represents that there is a first connection line segment between A and A i u(i) ​​​​there is a first connection line segment between A i pointing to A u(i) , GX iu(i) 1 = 2 indicates A i and A u(i) there is a first connection line segment between A u(i) pointing to A i , GX iu(i) 2 = p indicates A i and A u(i) there is a first connection line segment between A and the category of the first connection line segment is the pth first preset category, p = 1, 2, …, μ, μ is the total number of the first preset categories of the first connection line segment.

[0173] S012, if GX iu(i) 1 = GX iu(i) 2 = 0, it is determined that the data value of the ith row and the u(i)th column in ψ0is 0.

[0174] S013, if GX iu(i) 1 = 1 and GX iu(i) 2 = p, it is determined that the data value of the ith row and the u(i)th column in ψ0is p.

[0175] S014, if GX iu(i) 1 = 2 and GX iu(i) 2 = p, it is determined that the data value of the ith row and the u(i)th column in ψ0is -p.

[0176] S015, it is determined that the data value of the ith row and the ith column is 0.

[0177] S016, all the first human feature nodes are traversed, and the first node connection matrix ψ0is obtained.

[0178] wherein the data of the ith row and the u(i)th column in the first node connection matrix ψ0is the ith first human feature node A iThe data value corresponding to the connection relationship between the u(i)th first human body feature node, and the data value being 0 can represent that there is no direct association relationship between the corresponding one or two first human body feature nodes, the positive and negative of the data value can represent the direction of the arrow of the corresponding first connection line segment, and the absolute value size of the data value can represent the first preset category of the corresponding first connection line segment. Thus, the association relationship between the first human body feature nodes can be represented according to the first node connection matrix ψ0, and then the node feature of the target user can be provided as a feature basis for the material recommendation task of the target user.

[0179] In a specific embodiment, α = {α1, α2, …, α i , …, α m}, α i = {α i1 , α i2 , …, α ij , …, α in}, α ij represents the number of changes of the jth change category corresponding to the ith first human body feature node, and the data value of the ith row and the jth column in the first human body feature data matrix ψ1 is α ij .

[0180] The data value of the ith row and the jth column in the first human body feature data matrix ψ1 is the number of changes of the jth change category corresponding to the ith first human body feature node α ij .

[0181] In a specific embodiment, β = {β1, β2, …, β i , …, β m}, β i represents the node feature data corresponding to the ith second human body feature node, and the data value of the ith row in the second human body feature data matrix ψ2 is β i .

[0182] The data value of the ith row in the second human body feature data matrix ψ2 is the node feature data corresponding to the ith second human body feature node β i .

[0183] In a specific embodiment, γ = {γ1, γ2, …, γ i , …, γ m}, γ i represents the node feature data corresponding to the ith third human body feature node, and the data value of the ith row in the third human body feature data matrix ψ3 is γ i .

[0184] The data value of the ith row in the third human body feature data matrix ψ3 is the node feature data corresponding to the ith third human body feature node γi .

[0185] The above-mentioned third user graph and corresponding ψ0, ψ1, ψ2 and ψ3 are used to represent the abnormal situation of the target user in multiple dimensions, which provides a multi-dimensional data basis for the material recommendation task of the target user, thereby improving the accuracy of the material recommendation result.

[0186] S20, input the preset third user graph, ψ0, ψ1, ψ2 and ψ3 into the preset first encoding model, and obtain the target user embedding feature corresponding to the target user.

[0187] The first encoding model is used for feature extraction and encoding of the input third user graph, ψ0, ψ1, ψ2 and ψ3, and outputs the target user embedding feature corresponding to the target user, which represents the feature information of the first human body feature node, the second human body feature node and the third human body feature node of the target user, and provides multi-dimensional feature information for the material recommendation task of the target user, thereby improving the accuracy of the material recommendation result.

[0188] In a specific embodiment, the preset first encoding model is obtained by the following steps:

[0189] S021, obtain a plurality of sample users, a third user graph sample corresponding to each sample user, a first node connection matrix sample corresponding to each sample user, a first human body feature data matrix sample, a second human body feature data matrix sample and a third human body feature data matrix sample.

[0190] S022, input each third user graph sample, the first human body feature node connection matrix sample corresponding to each third user graph sample, the first human body feature data matrix sample, the second human body feature data matrix sample and the third human body feature data matrix sample into the first encoding model to be trained, and obtain the sample user embedding feature corresponding to each third user graph sample.

[0191] S023, input each sample user embedding feature into the preset first decoding model, and output a first node predicted connection matrix corresponding to each sample user.

[0192] S024, calculate the first model loss according to all first node connection matrix samples and all first node predicted connection matrices.

[0193] S025, train the parameters of the first encoding model to be trained according to the first model loss until the first model loss converges, and obtain the preset first encoding model.

[0194] The third user graph sample is obtained according to node connection information and node feature data of a sample user, and the third user graph sample, the first node connection matrix sample, the first human body feature data matrix sample, the second human body feature data matrix sample and the third human body feature data matrix sample are obtained by the same method as the third user graph, the first node connection matrix, the first human body feature data matrix, the second human body feature data matrix and the third human body feature data matrix.

[0195] The higher the similarity between the first node connection matrix sample and the corresponding first node predicted connection matrix, the higher the accuracy of the first encoding model. Therefore, the difference between each first node connection matrix sample and the corresponding first node predicted connection matrix is calculated, and the average of the difference between all first node connection matrix samples and all first node predicted connection matrices is determined as the first model loss, thereby improving the training accuracy of the first encoding model.

[0196] The above-mentioned first encoding model trained according to the third user graph sample, the first node connection matrix sample, the first human body feature data matrix sample, the second human body feature data matrix sample and the third human body feature data matrix sample has higher adaptability to the target user and higher accuracy, thereby improving the accuracy of the target user embedding feature.

[0197] S30, input the preset target user graph into the preset second encoding model to obtain the material embedding feature corresponding to the target user.

[0198] The target user graph represents the first human body feature node, the second human body feature node, the third human body feature node, the abnormal feature node and the material node, and the association relationship and the interaction relationship between the nodes. Compared with the third user graph, the target user graph represents the abnormal situation of the target user from more dimensions, thereby providing a data basis for material recommendation for the target user.

[0199] The second encoding model is used for feature extraction and encoding of the input target user graph, and outputs the material embedding feature corresponding to the target user to represent the association relationship between the material node and the first human body feature node, the second human body feature node, the third human body feature node and the abnormal feature node.

[0200] The above-mentioned target user graph is input into the preset second encoding model to obtain the material embedding feature corresponding to the target user, and the material recommendation task of the target user provides more dimensional feature information, thereby improving the accuracy of the material recommendation result.

[0201] S40, input the target user embedding feature and the material embedding feature into the preset prediction model to obtain a predicted connection value between the material node and the abnormal feature node.

[0202] The prediction model is used for feature coding and feature classification on the input target user embedding feature and the material embedding feature, and outputs the predicted connection value between the material node and the abnormal feature node, to represent the effective degree of the material corresponding to the material node in solving the abnormal situation corresponding to the abnormal feature node, as a basis for material recommendation for the target user.

[0203] The target user embedding feature and the material embedding feature are input into the preset prediction model to obtain the predicted connection value between the material node and the abnormal feature node, to represent the effective degree of the material corresponding to the material node in solving the abnormal situation corresponding to the abnormal feature node, for material recommendation for the target user, thereby improving the accuracy of the material recommendation result.

[0204] S50, determining the material node corresponding to the predicted connection value greater than the preset connection value threshold as a target material node, and recommending the material corresponding to the target material node to the target user.

[0205] The specific value of the preset connection value threshold can be set by the implementer according to the actual situation.

[0206] According to the preset third user graph, the first node connection matrix ψ0, the first human body feature data matrix ψ1, the second human body feature data matrix ψ2 and the third human body feature data matrix ψ3 corresponding to the target user are obtained, the preset third user graph, ψ0, ψ1, ψ2 and ψ3 are input into the preset first encoding model to obtain the target user embedding feature corresponding to the target user, to represent the feature information of the first human body feature node, the second human body feature node and the third human body feature node of the target user, the preset target user graph is input into the preset second encoding model to obtain the material embedding feature corresponding to the target user, to represent the association relationship between the material node and the first human body feature node, the second human body feature node, the third human body feature node and the abnormal feature node, and to provide multi-dimensional feature information for the material recommendation task of the target user; the target user embedding feature and the material embedding feature are input into the preset prediction model to obtain the predicted connection value between the material node and the abnormal feature node, the material node corresponding to the predicted connection value greater than the preset connection value threshold is determined as the target material node, and the material corresponding to the target material node is recommended to the target user, thereby improving the accuracy of the material recommendation result.

[0207] While certain specific embodiments of the application have been described in detail herein for the purposes of exemplification, numerous other variations and modifications will be apparent to persons skilled in the art. Alterations and modifications of detail can be made by those skilled in the art, having the benefit of the above description, without departing from the spirit and scope of the application. It is intended that all such alterations and modifications be included within the scope of the application whose limits are to be determined only by the appended claims.

Claims

1. A material recommendation method based on node screening, characterized in that, The node-based material recommendation method includes the following steps: S100: Obtain several first human feature nodes, second human feature nodes corresponding to each first human feature node, third human feature nodes corresponding to each first human feature node, several material nodes, several node pathways corresponding to all first human feature nodes, first human feature detection data corresponding to all first human feature nodes, first connection relationship data corresponding to each first human feature node, second human feature data corresponding to each second human feature node, and third human feature data corresponding to each third human feature node, as well as an abnormal feature node set E, wherein the first human feature node, second human feature node, and third human feature node are the target user's... When an abnormal situation occurs, different nodes will appear. The first human feature detection data is the whole exome sequencing data of the target user, which is used to characterize whether the first human feature corresponding to the first human feature node has changed. Each second human feature data is the information expression amount corresponding to the second human feature node. Each third human feature data is the information expression amount corresponding to the third human feature node. The substance corresponding to the substance node is the recommendation target recommended to the target user. The abnormal feature nodes in the abnormal feature node set E are used to characterize the abnormal situation of the target user. The substance node acts on the abnormal feature node and / or the first human feature node through the second human feature node to intervene in the abnormal situation. S200, based on the first human feature detection data, select the first abnormal human feature node from all the first human feature nodes, and determine the node path corresponding to the first abnormal human feature node as the first candidate node path. S300, based on the first connection relationship data, select the second abnormal human feature node from the first abnormal human feature node corresponding to each first candidate node path; S400, based on the positions of the second abnormal human feature node and the material node in the target user map, obtain the first node distance between each second abnormal human feature node and each material node; S500: Obtain the enrichment score of each node path, and select the second candidate node path from all node paths according to the preset first score threshold and the preset second score threshold. S600, based on the second human feature data corresponding to each second human feature node and the third human feature data corresponding to each third human feature node, obtain the target feature data corresponding to each first human feature node. S700, based on the target feature data, the preset first feature threshold and the preset second feature threshold, select the target human feature node from the first human feature node corresponding to each second candidate node path; S800, based on the positions of the target human body feature nodes and the material nodes in the target user map, obtain the second node distance between each target human body feature node and each material node; S900: Based on the distance between the first node, the distance between the second node, and a preset distance threshold, select the target material node from all material nodes.

2. The material recommendation method based on node screening according to claim 1, characterized in that, S200 also includes the following steps: S201, based on the first human feature detection data, determine the state detection category corresponding to each first human feature node, wherein the state detection category is abnormal or not abnormal; S202, when the state detection category corresponding to the first human feature node is abnormal, the corresponding first human feature node is determined as the first abnormal human feature node.

3. The material recommendation method based on node screening according to claim 1, characterized in that, The target user map is obtained through the following steps: S1, obtain the first human feature node set A, the first human feature dataset α, the second human feature dataset β, the third human feature dataset γ, the abnormal feature node set E, and the material node set F corresponding to the target user, as well as the second human feature node set B, the third human feature node set C, and the first connection relationship dataset G corresponding to the first human feature node set A, and the second connection relationship dataset H corresponding to the second human feature node set B, the fourth connection relationship dataset Y corresponding to the abnormal feature node set E, and the fifth connection relationship dataset Z corresponding to the material node set F, where A = {A1, A2, ..., A...} i , ..., A I }, A i It refers to the i-th first human feature node, B = {B1, B2, ..., B...} i , ..., B I }, B i Refers to A i The corresponding second human feature node, C={C1, C2, ..., C... i , ..., C I }, C i Refers to A i The corresponding third human feature node, G is used to represent the connection relationship between the first human feature nodes, H is used to represent the connection relationship between the second human feature nodes, Y is used to represent the connection relationship between the abnormal feature node and the first human feature node, the second human feature node and the third human feature node, Z is used to represent the connection relationship between the abnormal feature node and the second human feature node, i=1,2,...,I, u(i)=1,2,...,i-1,i+1,...,I, I refers to the total number of first human feature nodes, j=1,2,...,n, n refers to the total number of change categories; S2, based on the first human feature node set A, the second human feature node set B, and the third human feature node set C, obtain the first user graph set P = {P1, P2, ..., P...} i ..., P I }, where the i-th first user graph P i By analyzing A i B i and C i By connecting them pairwise; S3, based on the first connection relationship dataset G and the second connection relationship dataset H, connect all the first user graphs in the first user graph set P to obtain the second user graph; S4, based on the first human feature dataset α, the second human feature dataset β, and the third human feature dataset γ, update the node feature data of all the first human feature nodes, the second human feature nodes, and the third human feature nodes in the second user map, and obtain the third user map corresponding to the target user; S5, based on the fourth connection relationship dataset Y and the fifth connection relationship dataset Z, add all abnormal feature nodes in the abnormal feature node set E and all material nodes in the material node set F to the third user graph to obtain the target user graph.

4. The material recommendation method based on node screening according to claim 3, characterized in that, The first connection relation dataset G = {G1, G2, ..., G...} i , ..., G I }, G i ={G i1 G i2 , ..., G iu(i) , ..., G iI }, G iu(i) Refers to A i and A u(i) The first connection relationship data between them, which is used to indicate whether there is a first connecting line segment, the direction of the arrow of the first connecting line segment, and the first preset category of the first connecting line segment, S3 also includes the following steps: S311, obtain A i and A u(i) The first connection relationship data G between them iu(i) ={GX iu(i) 1 GX iu(i) 2 }, where GX iu(i) 1 =GX iu(i) 2 =0 indicates A i and A u(i) There is no first connecting line segment between them, GX iu(i) 1 =1 indicates A i and A u(i) There is a first connecting line segment, and the arrow points from A. i Pointing to A u(i) GX iu(i) 1 =2 represents A i and A u(i) There is a first connecting line segment, and the arrow points from A. u(i) Pointing to A i GX iu(i) 2 =ρ represents A i and A u(i) There exists a first connecting line segment and the category of the first connecting line segment is the ρth first preset category, where ρ = 1, 2, ..., μ, and μ refers to the total number of the first preset categories of the first connecting line segment; S312, if GX iu(i) 1 =GX iu(i) 2 =0, then A is not connected. i and A u(i) ; S313, if GX iu(i) 1 =1 and GX iu(i) 2 =ρ, then according to the arrow pointing from A i Pointing to A u(i) And the first connecting line segment A belongs to the first preset category of type ρ. i and A u(i) ; S314, if GX iu(i) 1 =2 and GX iu(i) 2 =ρ, then according to the arrow pointing from A u(i) Pointing to A i And the first connecting line segment A belongs to the first preset category of type ρ. i and A u(i) .

5. The material recommendation method based on node screening according to claim 4, characterized in that, The node pathway is obtained through the following steps: S101, for the i-th first human feature node A i If G i1 1 =G i2 1 =……=G iu(i) 1 =……=G iI 1 If A ≠ 1, then A i Updated to a first-level target node; S102, according to G i , will point to A i The first human feature node is updated to A i The corresponding second-level target node; S103, according to G, will point to A. i The first human feature node of the corresponding ω-th level target node is updated to A. i The corresponding target node at level ω+1, up to A i The total number of target nodes at level ω+1 is equal to 0, where ω is an integer greater than 1; S104, according to A i A i The corresponding second-level target node, A i The corresponding third-level target node to A i The order of the corresponding ω-th level target nodes is used to obtain A. i The corresponding node path; S105, traverse the first human feature node set A to obtain all node paths.

6. The material recommendation method based on node screening according to claim 5, characterized in that, The S300 also includes the following steps: S301, for any first candidate node path, obtain the level number corresponding to each first abnormal human feature node in the first candidate node path; S302, the first abnormal human feature node corresponding to the smallest level number in the first candidate node path is determined as the second abnormal human feature node corresponding to the first candidate node path.

7. The material recommendation method based on node screening according to claim 3, characterized in that, S400 also includes the following steps: S401, Obtain the t-th material node F t The ε-th second human feature node Z is connected t 1 ε and Z t 1 ε The first human feature node connected will be Z t 1 ε The first human feature node connected is determined to be F. t The corresponding first reference human body feature node; S402, for any second abnormal feature node, if the second abnormal feature node and F t If the corresponding first reference human feature node is on the same node path, then the first-level sequence number of the second abnormal feature node on the node path is obtained, along with F. t The corresponding first reference human feature node has a second-level sequence number on the node path; S403, based on the first level sequence number and the second level sequence number, obtain the second abnormal feature node and F. t The distance between the first nodes; S404, if the second abnormal feature node and F t If the corresponding first reference human feature nodes are not on the same node path, then the preset node distance is determined as the distance between the second abnormal feature node and F. t The distance between the first nodes.

8. The material recommendation method based on node screening according to claim 1, characterized in that, If the preset second score threshold is less than the preset first score threshold, S500 further includes the following steps: For any node path, if the enrichment score corresponding to the node path is greater than the preset first score threshold, or if the enrichment score corresponding to the node path is less than the preset second score threshold, then the node path is determined as the second candidate node path.

9. The material recommendation method based on node screening according to claim 1, characterized in that, If the preset second feature threshold is less than the preset first feature threshold, S700 further includes the following steps: For any target feature data, if the target feature data is greater than the preset first feature threshold, or if the target feature data is less than the preset second feature threshold, then the first human feature node corresponding to the target feature data is determined as the target human feature node.

10. The material recommendation method based on node screening according to claim 1, characterized in that, The S900 also includes the following steps: S901, based on the first node distance and the preset first distance threshold, select the first recommended material node from all material nodes and obtain the first recommended material node set; S902, based on the distance to the second node and the preset first distance threshold, select the second recommended material nodes from all material nodes and obtain the set of the second recommended material nodes; S903, the intersection of the first recommended material node set and the second recommended material node set is determined as the set corresponding to the target material node.

Images