CN109657122A - A kind of Academic Teams' important member's recognition methods based on academic big data - Google Patents
A kind of Academic Teams' important member's recognition methods based on academic big data Download PDFInfo
- Publication number
- CN109657122A CN109657122A CN201811500188.0A CN201811500188A CN109657122A CN 109657122 A CN109657122 A CN 109657122A CN 201811500188 A CN201811500188 A CN 201811500188A CN 109657122 A CN109657122 A CN 109657122A
- Authority
- CN
- China
- Prior art keywords
- scholar
- academic
- team
- node
- leader
- 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.)
- Granted
Links
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
Academic Teams' important member's recognition methods based on academic big data that the invention discloses a kind of, from the angle recognition team leader of degree centrality, the H factor and academic age, and use the DHRank model based on PageRank method for the first time to assess and mark team key members.In order to reasonably carry out team's cluster, the present invention is after carrying out team's cluster based on shortest path centered on leader, cluster result has been carried out using the method compared with leader's H factor is compared based on team size it is further perfect, help to improve the later period using DHRank model identification key members result accuracy.Compared to other methods, method proposed by the present invention is more accurate in terms of finding powerful scholar, and possesses lower operation time complexity.The present invention provides a kind of new methods that leader and key person are identified in Academic Teams, provide a kind of new solution for team's scientific research in academic network.
Description
Technical field
The invention belongs to academic Network Recognition technical field, be related to a kind of Academic Teams based on academic big data it is important at
Member's recognition methods, and in particular to a kind of degree centrality, the H factor and the recognition methods of the multiple features leader node at academic age and
Key members' identification is carried out based on the improved DHRank algorithm of PageRank.
Background technique
With the diversified development of current scientific research, Academic Teams are even studied to across university, research center
The trend development in field.The popularity and complexity of scientific research cooperative increase the scale of Academic Teams also increasingly, Team Member
Between and the also more sophisticated of the cooperative relationship between team and team.Team important member is the pillar of a team, they are not
The planning management responsibility inside team is only carry, the responsibility of whole team foreign exchanges cooperation is also carry.For in team
The research of important member can preferably reflect the cooperation architecture and development trend of team.However the growth of academic network size
The difficulty of team important member identification is increased with the complexity of cooperative relationship, while most researchs are also only conceived to team at present
And the identification of team leader, problem, which is not solved comprehensively and accurately also, to be identified for team important member.Therefore it needs to grind
Study carefully personnel and proposes that one kind can rationally be come for Academic Teams' structure and composition in accurate Academic Teams important member recognition methods
Relationship is preferably studied.
Summary of the invention
The present invention is directed to some shortcomings of existing research, proposes a kind of availability centrality, the H factor and scholar
The indexs combination DHRank algorithm such as academic age carries out the recognition methods of team leader and important member.This method is from macro
Network structure angle, the academic of scholar and the Efficiency angle of scholar is seen to identify the leader of team；
Team's building based on shortest path is carried out centered on team leader, uses the DHRank based on PageRank algorithm improvement
Algorithm carries out capability evaluation to the member in team, and obtains the key members in team for studying Academic Teams' composition and pre-
Survey following team leader.The present invention is the influential leader node and key members identified in Academic Teams
The problem of a kind of new solution is provided.
Technical solution of the present invention:
A kind of Academic Teams' important member's recognition methods based on academic big data, steps are as follows:
Step 1): the situation calculating target window time limit is quoted by counting true meeting paper, and according to target window
Scholar's paper in the time limit establishes academic network；
Step 2): the connection and scholar's paper of scholar's node in constructed academic network in step 1) are utilized
True reference amount carries out statistics calculating to degree centrality of the scholar in academic network and the scholar H factor, is potentially rolled into a ball
Team's leader's list；
Step 3): counting scholar's paper publishing situation in academic network and calculates the academic age of scholar, utilizes academic year
Obtained potential team leader's list determines leader node in age and step 2), and utilizes leader node and leader
Non- leader node is divided the Academic Teams for belonging to different leaders by the shortest path between person；
Step 4): the Academic Teams according to obtained in step 3), using DHRank algorithm in Academic Teams it is crucial at
Member is identified and is marked, and the important member in Academic Teams, including leader and Academic Teams key members are obtained.
Step 1)~step 4) is specific as follows:
Step 1):
By the cooperation of paper publishing situation and scholar in statistical analysis DBLP data set, determines and made with 10 years
The identification and analysis of team important member are carried out for the target window time limit.Paper collaboration, which is considered, cooperates to close between scholar
The specific manifestation of system, academic network based on this can really reflect the cooperation and team's ownership of scholar.Really
It sets the goal after the time limit, establishes academic network using scholar's paper information in the target time limit.Wherein, the node table in academic network
Show scholar, side indicates that the cooperative relationship between scholar, the weight on side indicate the cooperation paper number of two scholars.
Step 2) comprising the following three steps:
2.1) degree centrality of the scholar in academic network is calculated, according to other scholars in some scholar and academic network
Collaboration relationship and academic network in node number the degree centrality of the scholar is calculated；Degree centrality is network point
The most direct Measure Indexes of node significance level are portrayed in analysis, the bigger degree centrality for indicating this node of the degree of a node is more
Height, this node are more important in a network；It is important in a scientific cooperation team that scholar can be measured according to degree centrality
Property and influence power, it is as follows to spend central calculation formula:
Wherein, deg (a, G) indicates whole co-author's numbers of the scholar a in academic network G, and A indicates complete in academic network
Scholar's number in portion.
2.2) the H factor of scholar is calculated；The H factor is also known as H index, and ' H ' therein represents the high reference number of scholar, and one
The H factor of name scholar, which refers to the scholar at most, has H paper to be cited respectively at least H times.The H factor can be relatively accurately anti-
The scholastic attainment of a people is answered, the H index of scholar is higher, shows that the paper impact factor of the scholar is bigger.According to determining for the H factor
Justice, the true reference amount for counting paper publishing calculate the H factor of the scholar in academic network.
2.3) according to two indexs of the degree centrality of the obtained scholar of counting statistics and the H factor, it is equal that two indexs are found out
Higher than the scholar of index average value, and it is marked as potential team leader's node.
Step 3) includes following two step:
3.1) the academic age of scholar is calculated: according to every scholar's science in the available academic network of paper publishing situation
Age, i.e., from delivering time span of first paper to a last paper.The paper publishing behavior of scholar is counted,
The academic age, the scholar less than 5 was usually considered student researcher, these people can not be referred to as real team
Leader.On the basis of potential team leader's list that step 2) obtains, more than or equal to 5 of academic age is filtered out
Person is simultaneously marked as leader node.
3.2) for the non-leader node in academic network, calculate the node to whole leader nodes shortest path
Distance, the leader node with shortest shortest path distance are the leader node of node thus, are by this node division
In the team of corresponding leader node.After carrying out team's division, the case where belonging to multiple team there are some nodes, therefore need to be into
The further team of row divides；If there are more common node in Liang Ge team, i.e. common node number is greater than average group
Team's number of nodes, then need the H factor for comparing two team leader's nodes, the leader node of the H factor with higher is as two
The leader node of a team's whole node.
Step 4) includes following two step:
4.1) degree centrality of all members in academic network in team is calculated, is calculated then in conjunction with step 2) is middle
To the H factor calculate the sum of degree centrality and the H factor of each scholar, the academic of scholar is measured；
4.2) the DHRank value of all members in team is calculated, the calculation of DHRank is as follows:
Wherein, DHR (ai) indicate scholar aiDHRank value, PR (ai) it is scholar aiPageRank value, Deg (ai) indicate
Scholar aiDegree in academic network, hindex (ai) it is scholar aiH factor values.
Wherein scholar a in academic networkiPageRank value calculation it is as follows:
Wherein, E indicates the set on all sides in academic network；(ai,aj) indicate scholar's node aiAnd ajBetween company
Side, PR (aj) indicate scholar a in networkiCooperation scholar ajPageRank value,Indicate scholar's node ajDegree.
If P=(PR1,PR2,…PRn)TPageRank vector is tieed up for n, A is the corresponding transfer matrix of academic network, such as
Shown in lower:
Wherein, A (ai,aj) indicate from scholar's node aiTo scholar's node ajTransition probability,Indicate scholar's node ai's
Degree.
Using poweriteration method, specifying the iteration initial value of PageRank value is 1.Due in academic network
Collaboration relationship is bidirectional relationship, and does not consider the case where single author publishes thesis, the PageRank minimum value when being iterated
It is set as 0.By taking turns iterative solution more:
Pk=ATPk-1
Finally converge on | | Pk-Pk-1| | < ξ, i.e., the difference of iteration is less than threshold xi twice.Changing at the end of k expression iteration
Generation number.
In invention, threshold xi=10-5。
After obtaining the DHRank value of each member in team, Team Member is subjected to descending arrangement, team according to the value
Member's is key directly proportional to its DHRank value.
Beneficial effects of the present invention: degree centrality, the H factor and academic year of the present invention using scholar in academic network
Three scholar's features of age carry out leader's identification in Academic Teams, and propose one kind using the method based on shortest path
New team's construction method.For more acurrate, the effective key members found in team, the invention also provides be based on
PageRank improved DHRank algorithm identifies the key members in team.The experimental results showed that method proposed by the present invention
It is more accurate in terms of finding powerful scholar, and compared to other methods, the present invention has lower operation time multiple
Miscellaneous degree.The present invention provides a kind of new methods that leader and key person are identified in Academic Teams, for the team in academic network
Scientific research provides a kind of new solution.
Detailed description of the invention
Fig. 1 is the academic age distribution situation of scholars and with certain academic age in defined cooperation window
The distribution situation of scholar's number proportion.
Fig. 2 and Fig. 3 respectively indicates the degree distribution situation of the scholar in academic network and the H factor distribution situation of scholar.
Fig. 4 is the academic age distribution situation in academic network, can be determined according to this distribution situation and screen potential leader
The paper publishing year of person is limited to 5 years.
Fig. 5 is in team's identification process, between different team the case where existing common Team Member.
Fig. 6 be part team identification as a result, be marked out in figure be the team leader.
Specific embodiment
It to make the objectives, technical solutions, and advantages of the present invention clearer, below will be to specific embodiment party of the invention
Formula is described in further detail.
Present example provides a kind of Academic Teams' important member's recognition methods based on academic big data, this method packet
It includes:
Step 1: by the cooperation of paper publishing situation and scholar in statistical analysis DBLP data set, with scholar
Between true paper collaborate data and construct academic network.Node table in network shows scholar, and side indicates the cooperation between scholar
Relationship.The cooperation behavior between scholar is carried out with the method for mathematical statistics by the cooperation between statistical analysis scholar
Analyze the Finding Cooperative window time limit.
The paper data that the present invention chooses in DBLP data set are tested, and DBLP is that computer field has authoritative number
According to collection, it has included the relevant information of more than 400 ten thousand papers, it is desirable to from experiment valid data used are wherein obtained it is necessary to right
Data set carries out screening and filtering.DBLP data set is pre-processed, paper and the whole author of only one author are removed
Paper of the academic age less than 5.The present invention carries out team's building based on collaboration relationship, it is believed that common co-authored papers are cooperation
The concrete embodiment of relationship, so the paper of only one author can not reflect the true cooperation of author；It is academic simultaneously
Age is less less than experience of 5 scholar in sphere of learning and is typically considered student researcher, can not become one
The leader or key members of the bigger Academic Teams of scale.In defined cooperation window, the academic age of scholars
The distribution situation of distribution situation and scholar's number proportion with certain academic age is as shown in Figure 1.Fig. 4 is in academic network
Academic age distribution situation, can determine that screening potential leader be limited to 5 years in paper publishing year according to this distribution situation.
Step 2: the degree using the true reference amount of paper in academic network and data set to scholar in academic network
Centrality and the H factor carry out statistics calculating and generate potential team leader's list.
2.1) degree centrality of the scholar in academic network is calculated, only considers the cooperative relationship between scholar, according to some
Person is with the node number in the collaboration relationship of other scholars in academic network and academic network to the degree centrality of the scholar
It is calculated.It is as follows to spend central calculation formula:
Wherein, deg (a, G) indicates whole co-author's numbers of the scholar a in academic network G, and A indicates complete in academic network
Scholar's number in portion.
2.2) according to the true reference amount of paper publishing can the H factor to the scholar in academic network calculate.
2.3) according to two indexs of the degree centrality of the obtained scholar of counting statistics and the H factor, it is equal that two indexs are found out
Higher than the scholar of index average value, and it is marked as potential team leader's node, the experimental results showed that the degree of scholar is flat
Mean value is that the average value of 7.8181, the H factor is 3.2178, specific distribution situation such as Fig. 2, Fig. 3 institute for spending centrality and the H factor
Show.
Step 3:
The academic age of scholar in academic network is counted according to the paper publishing situation in data set, in step 2 gained
To potential leader node's list in determine real leader node.Base is used centered on resulting real leader node
Carry out team's building in team's clustering method of shortest path, and after the completion of Primary Construction, for there are common node compared with
More team carries out further team's building, finally obtains the final result of team's building.
3.1) according to the academic age of every scholar in the available academic network of paper publishing situation, by calculating this
Index, which can be found out, to be studied for many years in academic scientific research field and publishes thesis than more frequently scholar, this condition pair is utilized
What is filtered out in step 2 is further screened potential leader section.Wherein academic scholar of the age less than 5 is labeled as
Real leader node.
3.2) for the non-leader node in academic network, calculate the node to whole leader nodes shortest path
Distance is the leader node of node thus with leader node of the shortest path apart from minimum value, by this node in set
It is divided into the team of corresponding leader node.After carrying out team's division, the case where belonging to multiple team there are some nodes,
Further division processing is carried out in this case.If there are more common node, i.e. common node in Liang Ge team
Number is greater than average team's number of nodes, then needs the H factor for comparing two team leader's nodes, the neck of the H factor with higher
Leader node of the person's of leading node as Liang Ge team whole node.It is existing total between different team in team's identification process
The case where with Team Member, is as shown in Figure 5.
By above step, the leader of our available all Academic Teams and member's distribution.
Step 4:
In Academic Teams, not only only leader is managed whole team, also there are some key members
To assist leader to manage whole team.Therefore this step is using obtained team and based on PageRank algorithm
DHRank algorithm is identified and is marked to the key members in team.
4.1) calculate team in degree centrality of all members in academic network, then in conjunction with the H being calculated because
Son calculates the sum of degree centrality and H factor of each scholar, measures for the academic to scholar.
4.2) PageRank value of all members in team, scholar a are calculatediPageRank value calculation it is as follows:
Wherein, E indicates the set on all sides in academic network；(ai,aj) indicate scholar's node aiAnd ajBetween company
Side, PR (aj) indicate scholar a in networkiCooperation scholar ajPageRank value,Indicate scholar's node ajDegree.If P=
(PR1,PR2,…PRn)TPageRank vector is tieed up for n, A is the corresponding transfer matrix of academic network, as follows:
Wherein, A (ai,aj) indicate from scholar's node aiTo scholar's node ajTransition probability,Indicate scholar's node ai's
Degree.
Using poweriteration method, specifying the iteration initial value of PageRank value is 1.Due in academic network
Collaboration relationship is bidirectional relationship, and inventor does not consider the case where single author publishes thesis herein, when being iterated
PageRank minimum value is set as 0.By taking turns iterative solution more:
Pk=ATPk-1
Finally converge on | | Pk-Pk-1| | < ξ, i.e., the difference of iteration is less than some threshold value twice.Threshold value in the present embodiment
ξ=10-5。
4.3) the DHRank value of all members in team is calculated, the calculation of DHRank is as follows:
Wherein, DHR (ai) indicate scholar aiDHRank value, PR (ai) it is scholar aiPageRank value, Deg (ai) indicate
The scholar aiNode degree in academic network, hindex (ai) be the scholar H factor values.
4.4) after the DHRank value for obtaining each member in team, Team Member is subjected to descending arrangement according to the value,
Team Member it is key directly proportional to its DHRank value.Key members in team can help researcher clear in some group
In team, over time, the leader in team may be replaced by some key members.Fig. 6 is the identification of part team
As a result, be marked out in figure be the team leader.
Claims (2)
1. a kind of Academic Teams' important member's recognition methods based on academic big data, which is characterized in that steps are as follows:
Step 1): the situation calculating target window time limit is quoted by counting true meeting paper, and according to the target window time limit
Interior scholar's paper establishes academic network；
By statistical analysis DBLP data set in paper publishing situation and the cooperation between scholar, determine using 10 years as
The identification and analysis of target window time limit progress team important member；It recycles scholar's paper information in the target time limit to establish to learn
Art network；Node table in academic network shows scholar, and side indicates that the cooperative relationship between scholar, the weight on side indicate two scholars'
Cooperate paper number；
Step 2): the paper of the connection and scholar of scholar's node is true in the academic network constructed by utilizing in step 1)
Real reference amount carries out statistics calculating to degree centrality of the scholar in academic network and the scholar H factor, obtains potential team
Leader's list；
2.1) degree centrality of the scholar in academic network is calculated；According to the conjunction of other scholars in some scholar and academic network
Node number in work relationship and academic network calculates the degree centrality of the scholar；Spend central calculation formula such as
Under:
Wherein, deg (a, G) indicates whole co-author's numbers of the scholar a in academic network G, and A indicates whole in academic network
Scholar's number；
2.2) the H factor of scholar is calculated；The H factor of one scholar, which refers to the scholar at most, has H paper to be cited respectively extremely
It is H times few；According to the definition of the H factor, the true reference amount for counting paper publishing counts the H factor of the scholar in academic network
It calculates；
2.3) according to two indexs of the degree centrality of the obtained scholar of counting statistics and the H factor, two indexs is found out and are above
The scholar of index average value, and it is marked as potential team leader's node；
Step 3): counting scholar's paper publishing situation in academic network and calculates the academic age of scholar, using the academic age and
Obtained potential team leader's list determines leader node in step 2), and using leader node and leader it
Between shortest path non-leader node divided belong to the Academic Teams of different leaders；
3.1) it calculates the academic age of scholar: the paper publishing behavior of scholar being counted, is obtained in step 2) potential
On the basis of team leader's list, filters out scholar of the academic age more than or equal to 5 and be marked as leader node；
3.2) for the non-leader node in academic network, calculate the node to whole leader nodes shortest path away from
From the leader node with shortest shortest path distance is the leader node of node thus, is pair by this node division
It answers in the team of leader node；If common node number is greater than average team's number of nodes in Liang Ge team, compare two
The H factor of the leader node of team, leader of the leader node with the high H factor as Liang Ge team whole node
Node；
Step 4): the key members in team know using DHRank algorithm in the Academic Teams according to obtained in step 3)
It not and marks, obtains the important member in Academic Teams, including leader and team key members；
4.1) calculate degree centrality of all members in academic network in team, the H being calculated then in conjunction with step 2) because
Son calculates the sum of degree centrality and H factor of each scholar, measures to the academic of scholar；
4.2) the DHRank value of all members in team is calculated, the calculation of DHRank is as follows:
Wherein, DHR (ai) indicate scholar aiDHRank value, PR (ai) it is scholar aiPageRank value, Deg (ai) indicate scholar
aiDegree in academic network, hindex (ai) it is scholar aiH factor values；
Wherein, scholar a in academic networkiPageRank value calculation it is as follows:
Wherein, E indicates the set on all sides in academic network；(ai,aj) indicate scholar's node aiAnd ajBetween company side,
Indicate scholar's node ajDegree；
If P=(PR1,PR2,…PRn)TPageRank vector is tieed up for n, A is the corresponding transfer matrix of academic network, following institute
Show:
Wherein, A (ai,aj) indicate from scholar's node aiTo scholar's node ajTransition probability,Indicate scholar's node aiDegree；
Using poweriteration method, specifying the iteration initial value of PageRank value is 1, the PageRank when being iterated
Minimum value is set as 0, by taking turns iterative solution more:
Pk=ATPk-1
Finally converge on | | Pk-Pk-1| | < ξ, i.e., the difference of iteration is less than threshold xi twice；Iteration at the end of k indicates iteration is secondary
Number；
After obtaining the DHRank value of each member in team, Team Member is subjected to descending arrangement, Team Member according to the value
It is key directly proportional to its DHRank value.
2. a kind of Academic Teams' important member's recognition methods based on academic big data according to claim 1, feature
It is, threshold xi=10-5。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811500188.0A CN109657122B (en) | 2018-12-10 | 2018-12-10 | Academic team important member identification method based on academic big data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811500188.0A CN109657122B (en) | 2018-12-10 | 2018-12-10 | Academic team important member identification method based on academic big data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109657122A true CN109657122A (en) | 2019-04-19 |
| CN109657122B CN109657122B (en) | 2020-10-20 |
Family
ID=66113139
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811500188.0A Active CN109657122B (en) | 2018-12-10 | 2018-12-10 | Academic team important member identification method based on academic big data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109657122B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110851659A (en) * | 2019-10-23 | 2020-02-28 | 清华大学 | Student academic influence calculation method and system based on student thesis relationship network |
| CN110990662A (en) * | 2019-11-22 | 2020-04-10 | 北京市科学技术情报研究所 | Domain expert selection method based on citation network and scientific research cooperation network |
| CN112184051A (en) * | 2020-10-13 | 2021-01-05 | 中国工程物理研究院计算机应用研究所 | Employee work investigation method based on social network analysis technology |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103020302A (en) * | 2012-12-31 | 2013-04-03 | 中国科学院自动化研究所 | Academic core author excavation and related information extraction method and system based on complex network |
| CN103559262A (en) * | 2013-11-04 | 2014-02-05 | 北京邮电大学 | Community-based author and academic paper recommending system and recommending method |
| CN106886571A (en) * | 2017-01-18 | 2017-06-23 | 大连理工大学 | A kind of Forecasting Methodology of the scientific cooperation sustainability based on social network analysis |
| CN108564272A (en) * | 2018-04-08 | 2018-09-21 | 大连理工大学 | A kind of team's recommendation system building method based on Catfish Effect |
-
2018
- 2018-12-10 CN CN201811500188.0A patent/CN109657122B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103020302A (en) * | 2012-12-31 | 2013-04-03 | 中国科学院自动化研究所 | Academic core author excavation and related information extraction method and system based on complex network |
| CN103559262A (en) * | 2013-11-04 | 2014-02-05 | 北京邮电大学 | Community-based author and academic paper recommending system and recommending method |
| CN106886571A (en) * | 2017-01-18 | 2017-06-23 | 大连理工大学 | A kind of Forecasting Methodology of the scientific cooperation sustainability based on social network analysis |
| CN108564272A (en) * | 2018-04-08 | 2018-09-21 | 大连理工大学 | A kind of team's recommendation system building method based on Catfish Effect |
Non-Patent Citations (5)
| Title |
|---|
| FIZZA BIBI等: "Open AccessArticle Ranking Authors in an Academic Network Using Social Network Measures", 《MDPI》 * |
| YONGLI LI等: "A network-based and multi-parameter model for finding influential authors", 《JOURNAL OF INFORMETRICS》 * |
| 于永胜等: "基于社会网络分析的科研团队识别方法研究--基于迭代的中心度排名方法识别科研团队领导人", 《情报理论与实践》 * |
| 刘影: "复杂网络中节点影响力挖掘及其应用研究", 《中国博士学位论文全文数据库 基础科学辑》 * |
| 范如霞等: "基于合作网络的学者动态学术影响力模式识别研究", 《数据分析与知识发现》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110851659A (en) * | 2019-10-23 | 2020-02-28 | 清华大学 | Student academic influence calculation method and system based on student thesis relationship network |
| CN110990662A (en) * | 2019-11-22 | 2020-04-10 | 北京市科学技术情报研究所 | Domain expert selection method based on citation network and scientific research cooperation network |
| CN110990662B (en) * | 2019-11-22 | 2021-06-04 | 北京市科学技术情报研究所 | Domain expert selection method based on citation network and scientific research cooperation network |
| CN112184051A (en) * | 2020-10-13 | 2021-01-05 | 中国工程物理研究院计算机应用研究所 | Employee work investigation method based on social network analysis technology |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109657122B (en) | 2020-10-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Oyelade et al. | Application of k Means Clustering algorithm for prediction of Students Academic Performance | |
| CN104573359B (en) | A kind of mass-rent labeled data integration method of task based access control difficulty and mark person's ability | |
| CN104992184B (en) | A kind of multiclass image classification method based on semi-supervised extreme learning machine | |
| Alinezhad et al. | Sensitivity analysis of TOPSIS technique: the results of change in the weight of one attribute on the final ranking of alternatives | |
| Fang | Intelligent online English teaching system based on SVM algorithm and complex network | |
| Bulbul et al. | Comparison of classification techniques used in machine learning as applied on vocational guidance data | |
| Biswas et al. | Teaching and learning best differential evoltuion with self adaptation for real parameter optimization | |
| CN109657122A (en) | A kind of Academic Teams' important member's recognition methods based on academic big data | |
| Chen et al. | Identification of λ-fuzzy measures using sampling design and genetic algorithms | |
| Li et al. | Fuzzy-clustering embedded regression for predicting student academic performance | |
| CN107506350A (en) | A kind of method and apparatus of identification information | |
| CN108647800A (en) | A kind of online social network user missing attribute forecast method based on node insertion | |
| CN113361928B (en) | Crowd-sourced task recommendation method based on heterogram attention network | |
| CN108830407B (en) | Sensor distribution optimization method in structure health monitoring under multi-working condition | |
| CN107609084A (en) | One kind converges convergent resource correlation method based on gunz | |
| CN109508429A (en) | Personalized adaptive learning recommended method based on teaching platform big data analysis | |
| CN109840466A (en) | Based on the comprehensive multiple measurement learning method of cluster and the overall situation/local distance | |
| CN112148986A (en) | Crowdsourcing-based top-N service re-recommendation method and system | |
| Lian et al. | Cross-device user matching based on massive browse logs: The runner-up solution for the 2016 cikm cup | |
| CN112785156B (en) | Industrial collar and sleeve identification method based on clustering and comprehensive evaluation | |
| CN110188958A (en) | A kind of method that college entrance will intelligently makes a report on prediction recommendation | |
| CN113240209A (en) | Urban industry cluster development path prediction method based on graph neural network | |
| CN111292062B (en) | Network embedding-based crowd-sourced garbage worker detection method, system and storage medium | |
| CN111428861A (en) | Manufacturing enterprise overseas investment site selection decision method based on multi-scale multi-dimensional proximity | |
| JP2010079325A (en) | Model construction method, construction system, and program for construction |
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 |