CN108491409B - Urban medical system clustering method based on hospital associated network structural features - Google Patents

Urban medical system clustering method based on hospital associated network structural features Download PDF

Info

Publication number
CN108491409B
CN108491409B CN201810082241.3A CN201810082241A CN108491409B CN 108491409 B CN108491409 B CN 108491409B CN 201810082241 A CN201810082241 A CN 201810082241A CN 108491409 B CN108491409 B CN 108491409B
Authority
CN
China
Prior art keywords
hospital
network
node
doctor
representing
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201810082241.3A
Other languages
Chinese (zh)
Other versions
CN108491409A (en
Inventor
宣琦
李永苗
虞烨炜
郑钧
傅晨波
许荣华
阮中远
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhejiang University of Technology ZJUT
Original Assignee
Zhejiang University of Technology ZJUT
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Zhejiang University of Technology ZJUT filed Critical Zhejiang University of Technology ZJUT
Priority to CN201810082241.3A priority Critical patent/CN108491409B/en
Publication of CN108491409A publication Critical patent/CN108491409A/en
Application granted granted Critical
Publication of CN108491409B publication Critical patent/CN108491409B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients

Abstract

A method for clustering an urban medical system based on structural characteristics of a hospital association network comprises the following steps: step 1: collecting behavioral data regarding a doctor's multi-point practice; step 2: constructing a hospital-doctor bipartite network according to the multi-point practice behavior data of doctors; and step 3: projecting a hospital-doctor bipartite network to a hospital associated network of a single vertex; and 4, step 4: extracting structural features of the hospital associated network; and 5: and clustering the hospital association network by adopting a K-means algorithm. According to the invention, by constructing the hospital-doctor bipartite network, the projection of the bipartite network to the hospital associated network with a single vertex is realized, the structural characteristics of the hospital associated network are extracted, and the clustering of the hospital associated networks of the first-line city and the new first-line city can be realized by adopting a clustering algorithm. And the subsequent analysis of the distribution condition of the hospital network in each city by combining with policies, traffic conditions and the like implemented in each city can be realized.

Description

Urban medical system clustering method based on hospital associated network structural features
Technical Field
The invention relates to data mining, network science and machine learning technologies, in particular to a clustering method of an urban medical system based on hospital associated network structural features.
Background
Under the planned economic system, the medical health system is clearly positioned, Chinese medical health creates a series of glorious and achieves great achievements in various aspects of medical service, prevention, health care and the like. The medical service of rural area and town is also fully expanded at this moment, and the accessibility of medical service is greatly enhanced. Due to the influence of socioeconomic development and comprehensive national forces and the management of "political integration", china has problems in medical technology, service level and infrastructure construction to various degrees. This requires people to continuously explore new development approaches.
When talents get mobile or even leave free, how to attract more excellent talents or stay with current experts may become a subject to be placed in front of the courtyard. From the perspective of doctors, the biggest breakthrough of new medical improvement lies in the legalization of 'multi-point practice', in the future, doctors are not the national cadres any more, but become free operators, and return to the original attributes of the free employees, so that the method is favorable for greatly mobilizing the great support of most doctors on medical improvement, the enthusiasm and the support of the doctors are not available, and the new medical improvement is extremely difficult.
After the new medical improvement, the physician enters the sight of everybody again, and the new model of medical improvement is given the new worries. The development of medical complex construction is an important step and system innovation for deepening medical improvement, is favorable for adjusting and optimizing the structural layout of medical resources, promotes the downward movement of the center of gravity of medical and health work and the sinking of resources, promotes the basic service capacity, is favorable for the vertical communication of medical resources, promotes the overall efficiency of a medical service system, better implements graded diagnosis and treatment and meets the health requirements of the masses.
Disclosure of Invention
In order to overcome the defect that urban medical system clustering cannot be realized in the prior art, influence of doctor behaviors of first-line cities and new first-line cities on hospital associated networks is researched, and the influence degree of each city hospital associated network on policies is analyzed. The invention provides a clustering method of an urban medical system based on structural features of a hospital association network.
The technical scheme adopted by the invention for solving the technical problems is as follows:
a method for clustering an urban medical system based on structural characteristics of a hospital association network comprises the following steps:
step 1: collecting behavioral data regarding multiple points of practice by a doctor;
step 2: constructing a hospital-doctor bipartite network according to the multi-point practice behavior data of doctors;
and step 3: projecting a hospital-doctor bipartite network to a hospital associated network of a single vertex;
and 4, step 4: extracting structural features of the hospital associated network;
and 5: and clustering the hospital association network by adopting a K-means algorithm.
Further, in the step 1, behavior data about doctor's multi-point practice is collected, wherein the behavior data comprises doctor's name, hospital's name and hospital's address.
Further, in the step 2, according to the collected behavior data about the multipoint practice of the doctor in the step 1, a binary network is constructed, the binary network comprises two groups of different nodes which respectively represent the doctor and the hospital, and each group of nodes is not connected inside; the connecting edge of the two-division network indicates that the corresponding doctor performs the operation in the corresponding hospital.
Further, in the step 3, the bipartite network is projected onto the hospital nodes to form a hospital association network, which is a graph composed of sets V, E and W, and is denoted by G ═ V, E, W, where V ═ W1,v2,…,vNIs a set of nodes, each node representing a hospital;
Figure BDA0001561341620000021
is a set without directional connecting edges and represents the relationship among hospitals; w ═ Wij)N×NIs a set of side-by-side weights representing the number of people sharing a doctor in two hospitals. Projecting a bipartite network has two benefits: 1) information can be compressed; 2) the relationship among hospitals can be better revealed.
In the step 4, the network structure feature extraction includes the following processes:
4.1) the number of nodes | V | of the hospital association network, the node set of the hospital association network is:
V＝{v1,v2,…,vN}， (1)
wherein v isiIs a node of the network, and the node is a node,i is more than or equal to 1 and less than or equal to N, and the node set size is represented by | V | ═ N;
4.2) the number of connected edges | E | of the hospital association network, the set of edges of the hospital association network is:
Figure BDA0001561341620000039
wherein, the size of the edge set is represented by | E | ═ M;
4.3) centrality of the hospital association network: node viValue k ofiThe calculation formula of (a) is as follows:
Figure BDA0001561341620000031
where N represents the total number of nodes in the network,
Figure BDA0001561341620000032
node viIs of degree center DCiThe calculation is as follows:
Figure BDA0001561341620000033
wherein k isiRepresenting a node viA value of (d);
4.4) node v in Hospital's Association networkiMesomeric centrality BC ofiIs calculated as follows
Figure BDA0001561341620000034
Wherein, gstRepresenting a slave node vsTo node vtThe number of shortest paths of (a) to (b),
Figure BDA0001561341620000035
representing a slave node vsTo node vtG ofstStrip topPassing node v in short pathiThe number of shortest paths of (d);
4.5) feature vector centrality c of Hospital's associated networkse(G) The calculation is as follows:
Figure BDA0001561341620000036
wherein node v in the network graph GiCharacteristic vector centrality of (c)e(vi) The calculation is as follows:
Figure BDA0001561341620000037
where γ is a proportionality constant, aijThe elements of the weighted adjacency matrix representing the network,
Figure BDA0001561341620000038
hypothesis Ce＝(ce(v1),ce(v2),…,ce(vN))TIs the central vector of all nodes, equation (7) is written as the following matrix:
γCe＝ATCe， (8)
wherein A ═ aij)N×NRepresenting a weighted adjacency matrix, gamma being the corresponding eigenvalue, CeIs an adjacency matrix ATThe feature vector of (2);
4.6) Density D of Hospital-associated networks, calculated as follows:
Figure BDA0001561341620000041
wherein | E | represents the actual total number of edges of the network;
4.7) degree matching of the hospital association network r, calculated as follows:
Figure BDA0001561341620000042
wherein k isiAnd kjDegree of two end points of one continuous edge is represented, H represents total weight of all continuous edges in the network, and wijRepresenting a node viAnd node vjThe edge-connected weight of (1);
4.8) the clustering coefficient C of the hospital association network, calculated as follows:
Figure BDA0001561341620000043
wherein node v in network graph GiCluster coefficient of (C)iThe calculation is as follows:
Figure BDA0001561341620000044
si＝∑jwijrepresenting a node viIntensity of (k), kiRepresenting a node viThe degree of (a) to (b),
Figure BDA0001561341620000045
in the step 5, a K-means clustering algorithm is adopted to cluster the hospital association network, and the processing process is as follows: first, randomly selecting k samples, each sample initially representing a cluster center; for each sample remaining, assigning it to the nearest cluster according to its distance from the center of each cluster; and then recalculate each cluster center. This process is repeated until the criterion function converges.
The invention has the beneficial effects that: the urban medical system clustering method based on the structural features of the hospital association network realizes the projection of the bipartite network to the hospital association network of a single vertex by constructing the hospital-doctor bipartite network, simultaneously extracts the structural features of the hospital association network, and can realize the clustering of the hospital association networks of one-line and new-line cities by adopting a clustering algorithm. And the subsequent analysis of the distribution condition of the hospital associated network of the first-line city and the new first-line city in combination with policies, traffic conditions and the like implemented by each city can be realized.
Drawings
FIG. 1 is a flow chart of the programming of the present invention;
FIG. 2 is a diagram of a hospital-physician bipartite network according to the present invention;
fig. 3 is a diagram of a hospital association network according to the present invention.
Detailed Description
The following detailed description of embodiments of the invention is provided in connection with the accompanying drawings.
Referring to fig. 1 to 3, the invention relates to a method for clustering an urban medical system based on hospital associated network structure characteristics, wherein original data used by the method records information such as doctor names, hospital longitudes and latitudes and the like of multi-point practice. The invention extracts doctor names, doctor titles, hospital names and the like for multi-point practice.
The invention is divided into the following five steps:
step 1: collecting behavioral data regarding a doctor's multi-point practice;
step 2: constructing a hospital-doctor binary network according to the multi-point practice behavior data of the doctor;
and step 3: projecting a hospital-doctor bipartite network to a single-vertex hospital-associated network;
and 4, step 4: extracting structural features of the hospital associated network;
and 5: and clustering the hospital association network by adopting a K-means algorithm.
In the step 1, the specific operation process is as follows: collecting behavior data about doctor multi-point practice, wherein the behavior data comprises doctor names, hospital addresses and the like;
in the step 2, according to the collected behavior data about the multipoint practice of the doctor in the step 1, a hospital-doctor bipartite network is constructed, wherein the bipartite network comprises two groups of different nodes which respectively represent the doctor and the hospital, and each group of nodes is not connected inside; the connecting edge of the two-division network indicates that the corresponding doctor carries out the operation in the corresponding hospital.
In the step 3, projecting the bipartite network onto hospital nodes to form a hospital association network; the hospital association network is a graph composed of sets V, E and W, denoted as G ═ (V, E, W), where V ═ V1,v2,…,vNIs a set of nodes, each node representing a hospital;
Figure BDA0001561341620000053
is a set without directional connecting edges and represents the relationship among hospitals; w ═ Wij)N×NIs a set of side-by-side weights representing the number of people sharing a doctor in two hospitals. Projecting a bipartite network has two benefits: 1) information can be compressed; 2) the relationship among hospitals can be better revealed.
In the step 4, the network structure feature extraction includes the following processes:
4.1) the number of nodes | V | of the hospital association network, the node set of the hospital association network is:
V＝{v1,v2,…,vN}， (1)
wherein v isiIs a node, i is more than or equal to 1 and less than or equal to N, and the size of the node set is represented by | V | ═ N;
4.2) the number of connected edges | E | of the hospital association network, the set of edges of the hospital association network is:
Figure BDA0001561341620000051
wherein, the size of the edge set is represented by | E | ═ M;
4.3) centrality of the hospital association network: node viValue k ofiThe calculation formula of (a) is as follows:
Figure BDA0001561341620000052
where N represents the total number of nodes in the network,
Figure BDA0001561341620000061
node viIs of degree center DCiThe calculation is as follows:
Figure BDA0001561341620000062
wherein k isiRepresenting a node viA value of (d);
4.4) node v in Hospital-associated networkiMesomeric centrality BC ofiIs calculated as follows
Figure BDA0001561341620000063
Wherein, gstRepresenting a slave node vsTo node vtThe number of shortest paths of (a) to (b),
Figure BDA0001561341620000064
representing a slave node vsTo node vtG ofstPassing node v in shortest pathiThe number of shortest paths of (d);
4.5) feature vector centrality c of Hospital's associated networkse(G) The calculation is as follows:
Figure BDA0001561341620000065
wherein node v in network graph GiCharacteristic vector centrality of (c)e(vi) The calculation is as follows:
Figure BDA0001561341620000066
where γ is a proportionality constant, aijThe elements of the weighted adjacency matrix representing the network,
Figure BDA0001561341620000067
hypothesis Ce＝(ce(v1),ce(v2),…,ce(vN))TIs the central vector of all nodes, equation (7) is written as the following matrix:
γCe＝ATCe， (8)
wherein A ═ aij)N×NRepresenting a weighted adjacency matrix, gamma being the corresponding eigenvalue, CeIs an adjacency matrix ATThe feature vector of (2);
4.6) Density D of Hospital-associated networks, calculated as follows:
Figure BDA0001561341620000068
wherein | E | represents the actual total number of edges of the network;
4.7) degree matching of the hospital association network r, calculated as follows:
Figure BDA0001561341620000069
wherein k isiAnd kjDegree of two end points of one continuous edge is represented, H represents total weight of all continuous edges in the network, and wijRepresenting a node viAnd node vjThe edge-connected weight of (1);
4.8) the clustering coefficient C of the hospital association network, calculated as follows:
Figure BDA0001561341620000071
wherein node v in network graph GiCluster coefficient of (C)iThe calculation is as follows:
Figure BDA0001561341620000072
si＝∑jwijrepresenting a node viIntensity of (k) kiRepresenting a node viThe degree of (a) is greater than (b),
Figure BDA0001561341620000073
in the step 5, a K-means clustering algorithm is adopted to cluster the hospital association network, and the processing process is as follows: first, randomly selecting k samples, each sample initially representing a cluster center; for each sample remaining, assigning it to the nearest cluster according to its distance from the center of each cluster; then recalculate each cluster center. This process is repeated until the criterion function converges.
The invention uses the contour coefficient (silouette coefficient) to evaluate the clustering effect. As shown in table 1, when k is 2, the contour factor is the highest, and thus the hospital association network of the present invention is divided into 2 clusters. The final clustering results are shown in table 2.
k value silhouette coefficient
2 0.401
3 0.368
4 0.306
5 0.306
6 0.293
7 0.252
8 0.201
TABLE 1
Figure BDA0001561341620000074
Figure BDA0001561341620000081
TABLE 2
The invention extracts the structural characteristics of the hospital association network by constructing the hospital-doctor bipartite network and then projecting the data to the hospital nodes by using the collected data. The present invention is to be considered as illustrative and not restrictive. It will be understood by those skilled in the art that various changes, modifications and equivalents may be made therein without departing from the spirit and scope of the invention as defined in the appended claims.

Claims (4)

1. A method for clustering an urban medical system based on structural characteristics of a hospital association network is characterized by comprising the following steps:
step 1: collecting behavioral data regarding a doctor's multi-point practice;
step 2: constructing a hospital-doctor bipartite network according to the multi-point practice behavior data of doctors;
and step 3: projecting a hospital-doctor bipartite network to a hospital associated network of a single vertex;
and 4, step 4: extracting structural features of the hospital associated network;
and 5: clustering the hospital association network by adopting a K-means algorithm;
in the step 3, the bipartite network is projected to a hospital node to form a hospital association network; the hospital association network is a graph composed of sets V, E and W, denoted as G ═ (V, E, W), where V ═ V1,v2,…,vNIs a set of nodes, each node representing a hospital;
Figure FDA0003512422500000011
is a set of undirected continuous edges, which represents the relationship between hospitals; w ═ W (W)ij)N×NIs a continuous edge weight set which represents the number of people sharing doctors in two hospitals;
in the step 4, the network structure feature extraction includes the following processes:
4.1) the number of nodes | V | of the hospital association network, the node set of the hospital association network is:
V＝{v1,v2,…,vN}， (1)
wherein v isiIs a node, i is more than or equal to 1 and less than or equal to N, and the size of the node set is represented by | V | ═ N;
4.2) the number of connected edges | E | of the hospital association network, the set of edges of the hospital association network is:
Figure FDA0003512422500000012
wherein, the size of the edge set is represented by | E | ═ M;
4.3) centrality of hospital association network: node viValue k ofiThe calculation formula of (c) is as follows:
Figure FDA0003512422500000013
where N represents the total number of nodes in the network,
Figure FDA0003512422500000014
node viIs of degree center DCiThe calculation is as follows:
Figure FDA0003512422500000015
wherein k isiRepresenting a node viA value of (d);
4.4) node v in Hospital-associated networkiMesomeric centrality BC ofiIs calculated as follows
Figure FDA0003512422500000016
Wherein, gstRepresenting a slave node vsTo node vtThe number of shortest paths of (a) to (b),
Figure FDA0003512422500000021
representing a slave node vsTo node vtG ofstPassing node v in shortest pathiThe number of shortest paths of (d);
4.5) feature vector centrality c of Hospital's associated networkse(G) The calculation is as follows:
Figure FDA0003512422500000022
wherein node v in network graph GiCharacteristic vector centrality of (c)e(vi) The calculation is as follows:
Figure FDA0003512422500000023
where γ is a proportionality constant, aijThe elements of the weighted adjacency matrix representing the network,
Figure FDA0003512422500000024
hypothesis Ce＝(ce(v1),ce(v2),…,ce(vN))TIs the central vector of all nodes, equation (7) is written as the following matrix:
γCe＝ATCe， (8)
wherein A ═ aij)N×NRepresenting a weighted adjacency matrix, gamma being the corresponding eigenvalue, CeIs an adjacency matrix ATThe feature vector of (2);
4.6) Density D of Hospital-associated networks, calculated as follows:
Figure FDA0003512422500000025
wherein | E | represents the actual total number of edges of the network;
4.7) degree matching of the hospital association network r, calculated as follows:
Figure FDA0003512422500000026
wherein k isiAnd kjDegree of two end points of one continuous edge is represented, H represents total weight of all continuous edges in the network, and wijRepresenting a node viAnd node vjThe edge-connected weight of (1);
4.8) the clustering coefficient C of the hospital association network, calculated as follows:
Figure FDA0003512422500000027
wherein node v in network graph GiCluster coefficient of (C)iThe calculation is as follows:
Figure FDA0003512422500000028
si＝∑jwijrepresenting a node viIntensity of (k) kiRepresenting a node viThe degree of (a) to (b),
Figure FDA0003512422500000031
2. the urban medical system clustering method based on the hospital associated network structure characteristics according to claim 1, characterized in that: in the step 1, behavior data about multiple point practice of doctors is collected, and the behavior data comprises doctor names, hospital names and hospital addresses.
3. The urban medical system clustering method based on the hospital associated network structure characteristics according to claim 2, characterized in that: in the step 2, according to the collected behavior data about the multipoint practice of the doctor in the step 1, a bipartite network is constructed, wherein the bipartite network comprises two groups of different nodes which respectively represent the doctor and the hospital, and each group of nodes is not connected internally; the connecting edge of the two-division network indicates that the corresponding doctor performs the operation in the corresponding hospital.
4. The urban medical system clustering method based on the hospital associated network structure characteristics as claimed in one of claims 1 to 3, wherein: in the step 5, a K-means clustering algorithm is adopted to cluster the hospital association network, and the processing process is as follows: first, randomly selecting k samples, each sample initially representing a cluster center; for each sample remaining, assigning it to the nearest cluster according to its distance from the center of each cluster; then recalculating each cluster center; this process is repeated until the criterion function converges.
CN201810082241.3A 2018-01-29 2018-01-29 Urban medical system clustering method based on hospital associated network structural features Active CN108491409B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810082241.3A CN108491409B (en) 2018-01-29 2018-01-29 Urban medical system clustering method based on hospital associated network structural features

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810082241.3A CN108491409B (en) 2018-01-29 2018-01-29 Urban medical system clustering method based on hospital associated network structural features

Publications (2)

Publication Number Publication Date
CN108491409A CN108491409A (en) 2018-09-04
CN108491409B true CN108491409B (en) 2022-06-17

Family

ID=63343873

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810082241.3A Active CN108491409B (en) 2018-01-29 2018-01-29 Urban medical system clustering method based on hospital associated network structural features

Country Status (1)

Country Link
CN (1) CN108491409B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111310340B (en) * 2020-02-19 2022-08-16 中南大学 Urban area interaction abnormal relation identification method and equipment based on human movement

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106227835A (en) * 2016-07-25 2016-12-14 中南大学 Team's research direction method for digging based on two subnetwork figure hierarchical clusterings
CN106776781A (en) * 2016-11-11 2017-05-31 深圳云天励飞技术有限公司 A kind of human relation network analysis method and device

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106227835A (en) * 2016-07-25 2016-12-14 中南大学 Team's research direction method for digging based on two subnetwork figure hierarchical clusterings
CN106776781A (en) * 2016-11-11 2017-05-31 深圳云天励飞技术有限公司 A kind of human relation network analysis method and device

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
基于二分网络的网络用户评论情感分析;卢伟聪;《情报理论与实践》;20171130;第41卷(第2期);第121-126页 *

Also Published As

Publication number Publication date
CN108491409A (en) 2018-09-04

Similar Documents

Publication Publication Date Title
CN106780263B (en) High-risk personnel analysis and identification method based on big data platform
CN107430704A (en) Neural network algorithm is realized in nerve synapse substrate based on the metadata associated with neural network algorithm
CN105931116A (en) Automated credit scoring system and method based on depth learning mechanism
CN104809408A (en) Histogram release method based on difference privacy
CN108520365A (en) Education decision system based on big data analysis
CN105631749A (en) User portrait calculation method based on statistical data
CN112085250B (en) Population space distribution simulation method, terminal equipment and storage medium
CN111144693A (en) Decision-making method and device for urban public toilet site selection and computer readable storage medium
CN110297875A (en) It is a kind of to assess the method and apparatus that demand tightness is contacted between each functional areas in city
CN110290467A (en) The acquisition methods and device of dwell point, shopping centre service range, influence factor
Valerio et al. Accuracy vs. traffic trade-off of learning iot data patterns at the edge with hypothesis transfer learning
CN108491409B (en) Urban medical system clustering method based on hospital associated network structural features
CN104598648A (en) Interactive gender identification method and device for microblog user
CN109121093A (en) A kind of user's portrait construction method and system based on passive type WiFi and depth cluster
CN107451617A (en) One kind figure transduction semisupervised classification method
CN108876093A (en) A kind of many many wound design objective method for pushing created under platform
Xavier et al. Combining facility-location approaches for public schools expansion
CN117236465A (en) Information entropy-based federal decision tree information measurement method
CN109033746B (en) Protein compound identification method based on node vector
CN106485054A (en) Intelligent diagnostics data analysing method based on BP neural network algorithm and system
CN109657950A (en) Hierarchy Analysis Method, device, equipment and computer readable storage medium
CN106954207B (en) A kind of method and device for the account attributes value for obtaining target terminal
KR102241221B1 (en) Apparatus and method subdividing regional spaces of interest
CN114003803A (en) Method and system for discovering media account in specific region on social platform
CN107093149A (en) Online friend relation strength assessment method and system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant