CN110443574A - Entry convolutional neural networks evaluation expert's recommended method - Google Patents
Entry convolutional neural networks evaluation expert's recommended method Download PDFInfo
- Publication number
- CN110443574A CN110443574A CN201910677223.4A CN201910677223A CN110443574A CN 110443574 A CN110443574 A CN 110443574A CN 201910677223 A CN201910677223 A CN 201910677223A CN 110443574 A CN110443574 A CN 110443574A
- Authority
- CN
- China
- Prior art keywords
- expert
- matrix
- data
- project information
- information data
- 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
- 238000011156 evaluation Methods 0.000 title claims abstract description 32
- 238000000034 method Methods 0.000 title claims abstract description 31
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 24
- 239000011159 matrix material Substances 0.000 claims abstract description 68
- 238000013507 mapping Methods 0.000 claims description 7
- 230000004927 fusion Effects 0.000 claims description 4
- 238000013480 data collection Methods 0.000 claims description 3
- 238000012512 characterization method Methods 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 8
- 238000010276 construction Methods 0.000 abstract description 4
- 238000012545 processing Methods 0.000 abstract description 2
- 238000005516 engineering process Methods 0.000 description 8
- 230000006870 function Effects 0.000 description 7
- 230000004913 activation Effects 0.000 description 6
- 210000002569 neuron Anatomy 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 238000011176 pooling Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 238000007476 Maximum Likelihood Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000003339 best practice Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000012552 review Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/28—Databases characterised by their database models, e.g. relational or object models
- G06F16/284—Relational databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/10—Office automation; Time management
- G06Q10/103—Workflow collaboration or project management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- General Engineering & Computer Science (AREA)
- Marketing (AREA)
- Biomedical Technology (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Business, Economics & Management (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention relates to entry convolutional neural networks evaluation expert's recommended methods, belong to technical field of data processing.The present invention is respectively to existing a variety of incidence relations are analyzed between project information data and between expert's project information data, multiple project information data contents and several expert info data contents relevant to project are expressed as vector, and passes through the incidence relation between non-directed graph Construction of A Model project information data between expert info data and connects into matrix with its content vector respectively.Then convolutional neural networks are used, establish evaluation expert's recommended models of convergence project information data incidence relation and expert info data correlation relation, the model can learn the score relationship between project information data between expert info data, for evaluation expert's recommendation, which achieves preferable effect.
Description
Technical field
The present invention relates to entry convolutional neural networks evaluation expert's recommended methods, belong to technical field of data processing.
Background technique
In recent years, the activity such as declare, set up the project increasing, China's science and technology items at different levels and the scientific and technological people of China's science and technology item
Just management also realizes information system successively, and expert plays always crucial effect in evaluation.In order to guarantee
Objectivity, fairness and the fairness of project appraisal work, the work of selecting of evaluation expert are particularly important.Expert recommends
Be based on information of expert database abundant, found using technologies such as data mining, machine learning match with task to be recommended it is special
Family.From task to be recommended, expert recommends to can be regarded as one of personalized recommendation of Cempetency-based education special form
Formula.In e-commerce field, the research of proposed algorithm and technology is relatively more, this recommends to provide adequately theoretical base for expert
Plinth.Currently, mainly proposing following several method, 1. methods based on content in recommended method research;2. based on collaboration
The method of filtering;3. the method based on deep learning.
How to declare, set up the project etc. in activity in a large amount of science and technology item and selects satisfactory evaluation automatically fair and justly
Expert always is the key of project appraisal work, such as provides multiple project appraisal work, it will be able to directly select these projects
The evaluation expert of evaluation has very important application prospect.In expert's recommendation problem for entry evaluation, no
It is only Project evaluation content and the matched compactness of expert, shadow is similarly understood in existing various social relationships between expert
The determination of final review expert person is rung, and this is also to consider less factor in previous method simultaneously.Therefore, utilization is multiple
To the incidence relation between the content of Project evaluation, the content and expert of multiple experts, by artificial intelligence technology, how needle
Recommend evaluation expert as one of the difficult point of task and key technology multiple projects automatically.
Summary of the invention
The present invention provides entry convolutional neural networks evaluation expert's recommended methods, and this method is in multiple project datas
Hold and several expert data content representations relevant to project at vector, and by non-directed graph Construction of A Model project information data it
Between incidence relation between expert info data and connect into matrix with its content vector respectively;Then convolutional Neural net is used
Network establishes evaluation expert's recommended models of convergence project information data incidence relation and expert info data correlation relation, to answer
Recommend for evaluation expert.
The technical scheme is that entry convolutional neural networks evaluation expert's recommended method, the method it is specific
Steps are as follows:
Step1, data collection and vectorization indicate: assembled item and expert info data, by contents of a project information and specially
Family's content information is indicated with vector;
As a preferred solution of the present invention, in the step Step1, collection obtains the content number of project and expert architecture
According to, by one-hot coding by contents of a project information and expert's content information characterization become computer can recognize and handle to
Amount indicates, and indicates to be mapped to as intensive hidden vector expression by vector by Embedding layers.
Step2, project information data relationship and the building of expert info data relationship: it extracts between project information data and special
Existing incidence relation and its incidence relation matrix is constructed between family's information data;
As a preferred solution of the present invention, the step Step2 was recommended using the nonoriented edge of Markov network to characterize
Correlation in journey between project information data between expert info data constructs Markov network by its correlation, obtains
To the incidence relation matrix between project information data between expert info data.
Step3, fusion incidence relation matrix: on the basis of step Step1, Step2, by project information data matrix and
Expert info data matrix is associated with relational matrix connection respectively;
As a preferred solution of the present invention, the hidden vector that the step Step3 obtains step Step1 is respectively perpendicular connection
At project information data matrix and expert info data matrix, then by obtained project information data matrix, expert info number
Project information data correlation relation matrix, the expert info data correlation relation matrix obtained respectively with step Step2 according to matrix
Horizontal connection is to have merged the project matrix of project information data correlation relation and merged expert info data correlation relation
Expert's matrix.
Step4, recommended models building: on the basis of step Step3, believed based on convolutional neural networks using convergence project
The matrix for ceasing data correlation relation and expert info data correlation relation constructs evaluation expert's recommended models.
As a preferred solution of the present invention, convolutional neural networks are selected in the step Step4, using having merged project letter
It ceases the project information data matrix of data correlation relation and has merged the expert info data square of expert info data correlation relation
Battle array carries out convolution sum pond as input, and the result obtained by convolution sum pond is attached and learns and exports final
Scores realize that the multi-expert of entry is recommended.
The beneficial effects of the present invention are:
The present invention is respectively to existing a variety of incidence relations between project information data and between expert's project information data
It is analyzed, multiple project information data contents and several expert info data contents relevant to project is expressed as vector,
And by the incidence relation between non-directed graph Construction of A Model project information data between expert info data and respectively and in it
Hold vector and connects into matrix.Then convolutional neural networks are used, convergence project information data incidence relation and expert info are established
Evaluation expert's recommended models of data correlation relation, the model can learn between project information data between expert info data
Score relationship, realize multiple projects several experts recommend.The experimental results showed that the recommended method achieves preferable effect
Fruit, compared to not considering that the method for project information data relationship and expert info data relationship has certain mention in recommendation effect
It rises.
Detailed description of the invention
Fig. 1 is evaluation expert's incidence relation figure proposed by the present invention based on Markov network;
Fig. 2 is project appraisal expert's recommended models proposed by the present invention based on convolutional neural networks;
Fig. 3 is that the entry evaluation expert proposed by the present invention based on convolutional neural networks recommends block diagram.
Specific embodiment
Embodiment 1: as shown in Figure 1-3, entry convolutional neural networks evaluation expert's recommended method, the method it is specific
Steps are as follows:
Step1, data collection and vectorization indicate: collection obtains the content-data of project and expert architecture, utilizes
One-hot encodes the sparse vector for being converted into binaryzation, and sparse vector is become intensive by Embedding layers of mapping
Hidden vector indicate.
Step2, project information data relationship and the building of expert info data relationship: it extracts between project information data and special
Existing incidence relation and its incidence relation matrix is constructed between family's information data;
And the incidence relation between project information data between expert info data is extracted, expert info data
Incidence relation feature it is as shown in table 1, the incidence relation feature of project information data is as shown in table 2.
Using incidence relation feature construction expert's Markov network between expert info data, expert info data are obtained
Between incidence relation matrix.In expert's Markov network, fusion expert's incidence relation is calculated between expert info data
Correlation.In order to merge the expert info data correlation relation of these types, these expert info data correlation relations are defined
It is characterized function hm(ei,ej), and weight λ is distributed for each characteristic functionm, it is special that these associations are merged using log-linear model
Sign.Assuming that the correlation Sim (e between expert info datai,ej) indicate, correlation calculations formula is as follows:
Wherein,It indicates that m-th of characteristic function, M indicate incidence relation feature
Sum, λmIndicate the weight of character pair function,Indicate all and node eiThe set on connected side.Characteristic function weight λm, adopt
It is estimated with maximum Likelihood.As Sim (ei,ej) value when being greater than given threshold value beta, then it is assumed that Zhuan Jiaxin
Cease data eiWith expert info data ejBe it is relevant, i.e., there are a nothings between the two nodes in expert's Markov network
Xiang Bian, as shown in Figure 1, the weight on side is Sim (ei,ej).Similarly, the incidence relation matrix between project information data is constructed.
1 expert info data correlation relation feature of table
2 project information data correlation relation feature of table
Step3, fusion incidence relation matrix: on the basis of step Step1, Step2, by project information data matrix and
Expert info data matrix is associated with relational matrix connection respectively;
As a preferred solution of the present invention, the hidden vector that the step Step3 obtains step Step1 is respectively perpendicular connection
At project information data matrix and expert info data matrix, then by obtained project information data matrix, expert info number
Project information data correlation relation matrix, the expert info data correlation relation matrix obtained respectively with step Step2 according to matrix
Horizontal connection is to have merged the project matrix of project information data correlation relation and merged expert info data correlation relation
Expert's matrix.
Specifically, the hidden vector obtained after Embedding to be respectively perpendicular to the project matrix for connecting into n × k size
With expert's matrix of m × l size, wherein n is the number of project, and k is the dimension of the hidden vector of project, and m is the number of expert,
L is the dimension of the hidden vector of expert.
Then by the project matrix of obtained n × k size, m × l size expert's matrix respectively with the n of this n project ×
The relational matrix of n size, this m project m × m size relational matrix horizontal connection be merged project information data correlation
The project matrix of relationship and the expert's matrix for having merged expert info data correlation relation.
Step4, recommended models building: on the basis of step Step3, believed based on convolutional neural networks using convergence project
The matrix for ceasing data correlation relation and expert info data correlation relation constructs evaluation expert's recommended models.
As a preferred solution of the present invention, in conjunction with computer technology, using convolutional neural networks convergence project information data
Relationship and expert info data correlation relation construct evaluation expert's recommended models, as shown in Fig. 2, convolutional neural networks expert pushes away
Recommending the realization of model, specific step is as follows:
Step4.1, using convolutional neural networks to the item association matrix for having merged project information data correlation relation and
The expert's confederate matrix for having merged expert info data correlation relation carries out convolution to extract feature.Enable xi∈RnExpression project
The n-dimensional vector of i-th of project or expert in confederate matrix or expert's confederate matrix, then the matrix comprising m expert may be expressed as:
Wherein,Indicate attended operation, convolution operation refers to using the sliding window that length is h from x1:mVector is intercepted to make
For the input of convolutional layer activation primitive, the output obtained after calculating activation primitive is feature to be extracted.Such as following formula
It is shown:
ci=f (Wxi:i+h-1+b)
Wherein, function f indicates activation primitive, and the present invention selects ReLU function as activation primitive.W and b indicates weight square
Battle array and bias unit are the parameter of neural network, can be by being trained to obtain to neural network.ciIt indicates according to activation primitive
F and input vector xi:i+h-1Calculated feature.
Above-mentioned formula describe length be h sliding window in position xi:i+h-1The operation carried out when upper, works as sliding window
When mouth slides into ending from the beginning of sentence, one group of input can be obtained, be expressed as (x1:h,x2:1+h,...,xm-h+1:m), and then can be with
One group of output is obtained according to the input, is expressed as (c1,c2,...,cm-h+1), the output of this group is known as Feature Mapping.
Step4.2, the Feature Mapping of previous step is handled by max-pooling layers, obtain final feature to
Amount.It specifically, is the maximum value extracted in Feature Mapping, the thought done so is the maximum thought in each Feature Mapping
Value represents most important feature in the mapping.In addition, max-pooling layers of a major advantage is can be by means of this
The operation of layer obtains the feature vector of a regular length, and the length of vector is originally inputted without being concerned about.
Step4.3, project information data matrix and expert info data matrix are passed through into the result that convolution sum pond obtains
It is attached and learns and export final scores.Specifically at MLP layers, use Relu function as activation primitive, and
Softmax is carried out to result to handle to obtain probability.If softmax layers of categorical measure, i.e. neuron number are k, then:
Wherein, z indicates that softmax layers of input vector, W indicate the parameter of softmax layer network, yiIndicate output layer the
The output valve of i neuron, the result of final output are that k dimensional vector indicates the other probability value of k type.The output layer root of the model
The quantity of neuron is determined according to the categorical measure of classification task.N expert is directed to the score size of same project in order to obtain
Probability value, need to export n corresponding scores using n softmax in MLP layers of the last layer.
In order to illustrate performance of the invention, selects and do not use project information data correlation relation and expert info data correlation
The model of relational matrix compares, and input layer does not use project information data correlation relation matrix and expert info data correlation
Relationship is denoted as M1 as the method for auxiliary input, and input layer uses project information data correlation relation and expert info data correlation
Relationship is denoted as M2 as the method for auxiliary input;
The above-mentioned result for obtaining convolution sum pond is attached and learns and export the final average NDCG of distinct methods
Value, compares and analyzes expert's recommendation results by NDCG value, verifies the best practice of expert's recommendation.
Table 3 be distinct methods average NDCG value (wherein K be expert or project information data content expression in every class is set
Attribute is in the dimension size after Embedding).
Table 3: the average NDCG value of distinct methods
K | M1 | M2 |
8 | 0.901 | 0.905 |
16 | 0.913 | 0.919 |
32 | 0.918 | 0.920 |
As can be seen from the above data, use expert info data correlation relation, project information data correlation relation as
The effect that the model of auxiliary input obtains is better than without using expert info data correlation relation, project information data correlation relation
Model as auxiliary input.The results show, the present invention achieve preferably in the NDCG index that recommender system sorts
Effect.
Above in conjunction with attached drawing, the embodiment of the present invention is explained in detail, but the present invention is not limited to above-mentioned
Embodiment within the knowledge of a person skilled in the art can also be before not departing from present inventive concept
Put that various changes can be made.
Claims (5)
1. entry convolutional neural networks evaluation expert's recommended method, it is characterised in that: specific step is as follows for the method:
Step1, data collection and vectorization indicate: assembled item and expert info data, will be in contents of a project information and expert
Hold information is indicated with vector;
Step2, project information data relationship and the building of expert info data relationship: extracting between project information data and expert's letter
It ceases existing incidence relation between data and constructs its incidence relation matrix;
Step3, fusion incidence relation matrix: on the basis of step Step1, Step2, by project information data matrix and expert
Information data matrix is associated with relational matrix connection respectively;
Step4, recommended models building: on the basis of step Step3, convergence project Information Number is used based on convolutional neural networks
Evaluation expert's recommended models are constructed according to incidence relation and the matrix of expert info data correlation relation.
2. entry convolutional neural networks evaluation expert's recommended method according to claim 1, it is characterised in that: the step
In rapid Step1, collection obtains the content-data of project and expert architecture, by one-hot coding by contents of a project information and
The vector that expert's content information characterization becomes computer and can recognize and handle indicates, and is indicated vector by Embedding layers
The hidden vector that mapping becomes intensive indicates.
3. entry convolutional neural networks evaluation expert's recommended method according to claim 1, it is characterised in that: the step
Rapid Step2 characterized using the nonoriented edge of Markov network in recommendation process between project information data and expert info data it
Between correlation, by its correlation construct Markov network, obtain between project information data between expert info data
Incidence relation matrix.
4. entry convolutional neural networks evaluation expert's recommended method according to claim 1, it is characterised in that: the step
The hidden vector that rapid Step3 obtains step Step1, which is respectively perpendicular, connects into project information data matrix and expert info data square
Battle array, the project information for then obtaining obtained project information data matrix, expert info data matrix with step Step2 respectively
Data correlation relation matrix, expert info data correlation relation matrix horizontal connection are to have merged project information data correlation relation
Project matrix and merged expert's matrix of expert info data correlation relation.
5. entry convolutional neural networks evaluation expert's recommended method according to claim 1, it is characterised in that: the step
Select convolutional neural networks in rapid Step4, using the project information data matrix for having merged project information data correlation relation and
The expert info data matrix for having merged expert info data correlation relation carries out convolution sum pond as input, will pass through convolution
The result obtained with pond is attached and learns and export final scores, realizes that the multi-expert of entry is recommended.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910677223.4A CN110443574B (en) | 2019-07-25 | 2019-07-25 | Recommendation method for multi-project convolutional neural network review experts |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910677223.4A CN110443574B (en) | 2019-07-25 | 2019-07-25 | Recommendation method for multi-project convolutional neural network review experts |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110443574A true CN110443574A (en) | 2019-11-12 |
CN110443574B CN110443574B (en) | 2023-04-07 |
Family
ID=68431577
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910677223.4A Active CN110443574B (en) | 2019-07-25 | 2019-07-25 | Recommendation method for multi-project convolutional neural network review experts |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110443574B (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111782797A (en) * | 2020-07-13 | 2020-10-16 | 贵州省科技信息中心 | Automatic matching method for scientific and technological project review experts and storage medium |
CN113254642A (en) * | 2021-05-28 | 2021-08-13 | 华斌 | E-government affair project evaluation expert group recommendation method based on multi-dimensional feature balance |
US11200591B2 (en) * | 2018-05-30 | 2021-12-14 | Paypal, Inc. | Electronic content based on neural networks |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103136267A (en) * | 2011-12-01 | 2013-06-05 | 腾讯科技(深圳)有限公司 | Expert mining method and device based on community |
CN103631859A (en) * | 2013-10-24 | 2014-03-12 | 杭州电子科技大学 | Intelligent review expert recommending method for science and technology projects |
CN103823896A (en) * | 2014-03-13 | 2014-05-28 | 蚌埠医学院 | Subject characteristic value algorithm and subject characteristic value algorithm-based project evaluation expert recommendation algorithm |
CN108804689A (en) * | 2018-06-14 | 2018-11-13 | 合肥工业大学 | The label recommendation method of the fusion hidden connection relation of user towards answer platform |
CN109241424A (en) * | 2018-08-29 | 2019-01-18 | 陕西师范大学 | A kind of recommended method |
CN109446413A (en) * | 2018-09-25 | 2019-03-08 | 上海交通大学 | Serializing recommended method based on item associations relationship |
-
2019
- 2019-07-25 CN CN201910677223.4A patent/CN110443574B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103136267A (en) * | 2011-12-01 | 2013-06-05 | 腾讯科技(深圳)有限公司 | Expert mining method and device based on community |
CN103631859A (en) * | 2013-10-24 | 2014-03-12 | 杭州电子科技大学 | Intelligent review expert recommending method for science and technology projects |
CN103823896A (en) * | 2014-03-13 | 2014-05-28 | 蚌埠医学院 | Subject characteristic value algorithm and subject characteristic value algorithm-based project evaluation expert recommendation algorithm |
CN108804689A (en) * | 2018-06-14 | 2018-11-13 | 合肥工业大学 | The label recommendation method of the fusion hidden connection relation of user towards answer platform |
CN109241424A (en) * | 2018-08-29 | 2019-01-18 | 陕西师范大学 | A kind of recommended method |
CN109446413A (en) * | 2018-09-25 | 2019-03-08 | 上海交通大学 | Serializing recommended method based on item associations relationship |
Non-Patent Citations (3)
Title |
---|
SHENGXIANG GAO等: ""A Method to Review Expert Recommendation Using Topic Relevance and Expert Relationship"", 《INTERNATIONAL JOURNAL OF COOPERATIVE INFORMATION SYSTEMS》 * |
XIANGNAN HE等: ""Neural Collaborative Filtering"", 《PROCEEDINGS OF THE 26TH INTERNATIONAL CONFERENCE ON WORLD WIDE WEB》 * |
余峰等: ""基于主题信息的项目评审专家推荐方法"", 《计算机工程》 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11200591B2 (en) * | 2018-05-30 | 2021-12-14 | Paypal, Inc. | Electronic content based on neural networks |
CN111782797A (en) * | 2020-07-13 | 2020-10-16 | 贵州省科技信息中心 | Automatic matching method for scientific and technological project review experts and storage medium |
CN113254642A (en) * | 2021-05-28 | 2021-08-13 | 华斌 | E-government affair project evaluation expert group recommendation method based on multi-dimensional feature balance |
Also Published As
Publication number | Publication date |
---|---|
CN110443574B (en) | 2023-04-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111523047B (en) | Multi-relation collaborative filtering algorithm based on graph neural network | |
CN111428147B (en) | Social recommendation method of heterogeneous graph volume network combining social and interest information | |
CN112529168B (en) | GCN-based attribute multilayer network representation learning method | |
CN109614614A (en) | A kind of BILSTM-CRF name of product recognition methods based on from attention | |
CN112364976B (en) | User preference prediction method based on session recommendation system | |
CN108920544A (en) | A kind of personalized position recommended method of knowledge based map | |
CN110443574A (en) | Entry convolutional neural networks evaluation expert's recommended method | |
CN109241255A (en) | A kind of intension recognizing method based on deep learning | |
CN107220365A (en) | Accurate commending system and method based on collaborative filtering and correlation rule parallel processing | |
CN111274398A (en) | Method and system for analyzing comment emotion of aspect-level user product | |
CN107391582B (en) | The information recommendation method of user preference similarity is calculated based on context ontology tree | |
CN109241366B (en) | Hybrid recommendation system and method based on multitask deep learning | |
CN111753207B (en) | Collaborative filtering method for neural map based on comments | |
CN111127146A (en) | Information recommendation method and system based on convolutional neural network and noise reduction self-encoder | |
CN110442618A (en) | Merge convolutional neural networks evaluation expert's recommended method of expert info incidence relation | |
CN106157156A (en) | A kind of cooperation recommending system based on communities of users | |
CN114693397A (en) | Multi-view multi-modal commodity recommendation method based on attention neural network | |
CN112016002A (en) | Mixed recommendation method integrating comment text level attention and time factors | |
CN109033294A (en) | A kind of mixed recommendation method incorporating content information | |
CN112396492A (en) | Conversation recommendation method based on graph attention network and bidirectional long-short term memory network | |
CN110781401A (en) | Top-n project recommendation method based on collaborative autoregressive flow | |
Sun et al. | A content-enhanced approach for cold-start problem in collaborative filtering | |
CN111104975A (en) | Credit assessment model based on breadth learning | |
CN113822419A (en) | Self-supervision graph representation learning operation method based on structural information | |
CN112612948A (en) | Deep reinforcement learning-based recommendation system construction method |
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 |