CN111401764B - Comprehensive evaluation method for satisfaction degree of library users based on CPN network model - Google Patents

Comprehensive evaluation method for satisfaction degree of library users based on CPN network model Download PDF

Info

Publication number
CN111401764B
CN111401764B CN202010212527.6A CN202010212527A CN111401764B CN 111401764 B CN111401764 B CN 111401764B CN 202010212527 A CN202010212527 A CN 202010212527A CN 111401764 B CN111401764 B CN 111401764B
Authority
CN
China
Prior art keywords
cpn
user satisfaction
network model
layer
library
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
CN202010212527.6A
Other languages
Chinese (zh)
Other versions
CN111401764A (en
Inventor
肖三霞
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.)
Fujian Jiangxia University
Original Assignee
Fujian Jiangxia University
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 Fujian Jiangxia University filed Critical Fujian Jiangxia University
Priority to CN202010212527.6A priority Critical patent/CN111401764B/en
Publication of CN111401764A publication Critical patent/CN111401764A/en
Application granted granted Critical
Publication of CN111401764B publication Critical patent/CN111401764B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • 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
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
    • 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
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • 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
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Tourism & Hospitality (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • General Health & Medical Sciences (AREA)
  • Software Systems (AREA)
  • Biomedical Technology (AREA)
  • General Engineering & Computer Science (AREA)
  • Molecular Biology (AREA)
  • Mathematical Physics (AREA)
  • Evolutionary Computation (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Development Economics (AREA)
  • Educational Administration (AREA)
  • Computing Systems (AREA)
  • Entrepreneurship & Innovation (AREA)
  • General Business, Economics & Management (AREA)
  • Artificial Intelligence (AREA)
  • Marketing (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Quality & Reliability (AREA)
  • Operations Research (AREA)
  • Game Theory and Decision Science (AREA)
  • Primary Health Care (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention relates to a library user satisfaction comprehensive evaluation method based on a CPN network model, which is suitable for being executed in computer equipment and comprises the following steps: setting user satisfaction index variables, and setting a plurality of user satisfaction index variables according to evaluation criteria of user satisfaction evaluation on the library; setting user satisfaction levels, wherein the user satisfaction levels comprise a plurality of levels from low to high; constructing a CPN network model, taking the number of user satisfaction index variables as the number of input layer neurons, taking the number of user satisfaction grades as the number of output layer neurons, selecting the number of competitive layer neurons according to the number of input layer neurons and the number of output layer neurons, and constructing the CPN network model; and acquiring a sample set consisting of the user satisfaction index variable observed value and the user satisfaction value, training the CPN model by using the sample set, and evaluating the user satisfaction of the library by using the trained CPN model.

Description

Comprehensive evaluation method for satisfaction degree of library users based on CPN network model
Technical Field
The invention relates to the field of libraries, in particular to a library user satisfaction comprehensive evaluation method based on a CPN network model.
Background
User satisfaction is an important index for measuring the quality of service of the library. The positioning of libraries is undergoing a transition from resource-centric to reader-centric, and there is a continuing need to improve the level of library service to the fullest extent possible in response to the needs of readers. Therefore, the user satisfaction degree instead of the past evaluation indexes such as the collection amount, the borrowing amount and the circulation rate becomes an important standard for checking the service quality of the library of colleges and universities. However, with the continuous development of the knowledge and information age, the demands of readers are changing, and a unified reader satisfaction evaluation index system of libraries in colleges and universities is not formed so far. In addition, whether the user is satisfied with the service of the library belongs to the subjective category, the measurement is not easy, and the satisfaction degree of readers in the library is comprehensively and objectively evaluated by combining the characteristics of the library through a scientific method.
Currently, developed countries abroad have established user satisfaction assessment tools for most industries and enterprises, such as the european customer satisfaction index and the us customer satisfaction index. In china, although there is currently some research on user satisfaction, there is still a gap compared to abroad: the service research level is lack of depth and breadth of service research; the research level of the user satisfaction, system construction, model design, method improvement and other aspects are lack of innovation; in the aspect of empirical research, the method is still in the shallow conceptual research stage, and further practical research is lacked.
Disclosure of Invention
In order to solve the above problems in the prior art, the present invention provides a method for comprehensively evaluating user satisfaction of a library based on a CPN network model, so as to try to solve or at least alleviate the above problems.
The technical scheme of the invention is as follows:
a comprehensive evaluation method for the satisfaction degree of library users based on a CPN network model is suitable for being executed in computer equipment and comprises the following steps:
setting user satisfaction index variables, and setting a plurality of user satisfaction index variables according to the evaluation standard of the user on the evaluation of the satisfaction of the user on the library;
setting a user satisfaction degree grade, wherein the user satisfaction degree grade comprises a plurality of grades from low to high;
constructing a CPN network model, taking the number of user satisfaction index variables as the number N of input layer neurons, taking the number of user satisfaction grades as the number M of output layer neurons, selecting the number Q of competitive layer neurons according to the number N of the input layer neurons and the number M of the output layer neurons, and constructing the CPN network model;
and acquiring a sample set consisting of the user satisfaction index variable observed value and the user satisfaction value, training the CPN model by using the sample set, and evaluating the user satisfaction of the library by using the trained CPN model.
Further, the user satisfaction index variable comprises a plurality of primary index variables, each primary index variableA plurality of secondary index variables are also arranged under the index variables; the first-level index variables comprise three variables of facility environment conditions, service quality and resource construction; wherein the facility environmental condition comprises service system stability x1Service system response speed x2Network security and smoothness x3And resource acquisition facility x4Four secondary index variables; the quality of service comprises service flow normalization x5Service staff specialty x6And service content personalization x7Three secondary index variables; the resource construction comprises information resource authority x8Information content timeliness x9Resource structure completeness x10And information push accuracy x11And four secondary index variables, wherein the total number of the secondary index variables is used as the neuron number N of the input layer, and the input vector of the input layer is generated through each secondary index variable.
Further, the user satisfaction levels sequentially include five levels from low to high, namely low, normal, high and high.
Further, the number Q of neurons in the contention layer is selected from a candidate set U, which is determined by the following formula:
Figure BDA0002423308670000031
wherein R (. cndot.) is a rounding function.
Further, when the input vector is transmitted to the input layer of the CPN network model, the calculation value on the input layer is processed
Figure BDA0002423308670000033
And (6) carrying out normalization processing.
Further, in the CPN network model, the network inputs obtained by Q neurons in the competition layer are:
Figure BDA0002423308670000032
wherein, i is 1, …,11, j is 1, …, Q, W is the connection weight from the input layer to the competition layer; when the model is not trained, the initial value of each element of W is a random value in [0,1 ];
the network outputs of the Q neurons in the competition layer are: k ═ K1,…,kQ]；
In the formula, the i (i ═ 1, …, Q) th neuron output value kiThe calculation is performed as follows:
Figure BDA0002423308670000041
further, in the output layer, the network input F obtained by the M neurons adopts the following algorithm:
Figure BDA0002423308670000042
wherein i is 1, …, Q, j is 1, …,5, and V is the connection weight from the contention layer to the output layer; when the model is not trained, the initial value of each element of V is [0,1]]A random value of; elements Y of a net output vector Y of M neurons in a competition layeriThe calculation was performed using the following formula:
Figure BDA0002423308670000043
further, the sample set comprises a training sample subset and a testing sample subset, the training sample subset is used for training the CPN network model, the testing sample subset is used for testing the CPN network model, and the sample capacity of the training sample subset is greater than that of the testing sample subset; when the CPN model is applied to evaluate the satisfaction degree of the library user, the actual user satisfaction degree index variable is used as the input value of the CPN model to be calculated, the CPN model output value representing the comprehensive evaluation value of the user satisfaction degree is obtained, the comprehensive evaluation of the satisfaction degree of the library user is realized, a sample formed by the current index observation value and the comprehensive evaluation value of the user satisfaction degree is contained in the training sample subset or the testing sample subset, and the CPN model is updated at the same time.
Further, when the CPN network model is trained by using the training sample subset, the correction strategy of the weight W from the input layer to the competitive layer is as follows:
calculating a new connection weight W' according to the current connection weight W and the input vector:
W′＝W+α(X-W)；
wherein 0< α <1 is learning efficiency;
w 'is normalized, and W is set to W', thereby correcting W.
Further, when the training sample subset is adopted to train the CPN network model, the modification strategy of the connection weight V from the competition layer to the output layer is as follows:
calculating a new connection weight V' according to the current connection weight V and the input vector:
V′＝V+α(Y-V)
wherein 0< α <1 is learning efficiency;
and carrying out normalization processing on V ', and enabling V to be equal to V', so as to realize the correction of V.
The invention has the following beneficial effects:
1. the method for comprehensively evaluating the user satisfaction of the library based on the CPN model is convenient to realize and low in cost, ensures the practicability, reliability and operability of the model, greatly improves the estimation accuracy of the user satisfaction, is convenient to analyze factors influencing the user satisfaction, guides the improvement measures of the library, and indicates the direction for the improvement of the follow-up user satisfaction.
2. In the subsequent system use process, each evaluation is taken as a sample to be incorporated into a training sample subset or a test sample subset, and the connection weight of the CPN network model is recalculated to realize the dynamic update of the sample set.
Drawings
Fig. 1 is an operation flowchart of a library user satisfaction comprehensive assessment method based on a CPN network model according to an embodiment of the present invention.
Fig. 2 is a flowchart of a library user satisfaction comprehensive assessment calculation method based on a CPN network model according to an embodiment of the present invention.
Detailed Description
The invention is described in detail below with reference to the figures and the specific embodiments.
Fig. 1 shows an operation flow of a library user satisfaction comprehensive assessment method based on a CPN network model. As shown in fig. 1, the method starts with step S101, obtaining a user satisfaction index variable, where the user satisfaction index variable includes a plurality of primary index variables, and each primary index variable is provided with a plurality of secondary index variables. In this embodiment, the primary index variables include 3 variables such as facility environmental conditions, quality of service, and resource construction. Wherein the facility environmental conditions include service system stability x1Service system response speed x2Network security and smoothness x3Resource acquisition convenience x4Waiting for 4 secondary index variables; the quality of service comprises a service flow normalization x5Service staff specialty x6Service content personalization x7Waiting for 3 secondary index variables; the resource construction comprises information resource authority x8Information content timeliness x9Resource structure completeness x10Accuracy x of information push11And 4 secondary index variables are equal. Therefore, the number of the secondary index variables is 11, that is, the number N of neurons in the input layer of the CPN network model is 11, so that the input vector of the CPN network model is:
X＝[x1,x2,x3,x4,x5,x6,x7,x8,x9,x10,x11]
subsequently, the process proceeds to step S102, where a user satisfaction level is set, where the user satisfaction level includes a plurality of levels from low to high in sequence. In this embodiment, the user satisfaction levels include 5 levels, low, normal, high, and the like, in order from low to high.
Subsequently, the flow proceeds to step S10And 3, constructing a CPN network model. In this embodiment, the number of input vector elements is 11, that is, the number N of CPN network model input layer neurons is 11; the number of output vector elements is 5, that is, the number M of output layer neurons of the CPN network model is 5, which is y1、y2、y3、y4、y5Namely, the CPN network model output vector is:
Y＝[y1,y2,y3,y4,y5]
for the number of competition layer neurons, Q, the set of recommended values for Q is calculated as follows:
Figure BDA0002423308670000071
where R (·) is a rounding function, obtained from N ═ 11, M ═ 5:
U＝[3,5,6,7,8,9,10,11,12,13,14]
in this embodiment, Q is selected to be 7.
Then, S104 is entered, a sample set is obtained, and the sample set is divided into a training sample subset and a test sample subset, and generally, the sample capacity of the training sample subset is made larger than the sample capacity of the test sample subset. In this embodiment, the sample capacity of the training sample subset accounts for 80% of the total sample capacity, and the sample capacity of the test sample subset accounts for 20% of the total sample capacity.
After the above steps are completed, step S105 is entered to train the CPN network model. And training the CPN according to the selected training sample subset, and solving the connection weight of the CPN model. To make the dimension uniform, the input value of each neuron of the input layer is firstly trained
Figure BDA0002423308670000072
According to each x in the input vectoriNormalization was performed as follows:
Figure BDA0002423308670000081
the connection weight W from the input layer to the competition layer is:
Figure BDA0002423308670000082
where i is 1, …,11, j is 1, …, and Q, and when the model is not trained, the initial value of each element W is a random value within [0,1 ]. The net inputs obtained for the Q neurons in the competition layer are:
E＝[e1,…,eQ]＝XW
the network outputs of the Q neurons in the competition layer are:
K＝[k1,…,kQ]
in the formula, the i (i ═ 1, …, Q) th neuron output value kiThe calculation is performed as follows:
Figure BDA0002423308670000083
that is, in the competition layer, the neuron which obtains the maximum input is in the excited state, the output value is 1, the other neurons are in the inhibited state, and the output value is 0.
The connection weight V from the contention layer to the output layer is:
Figure BDA0002423308670000084
where i is 1, …, Q, j is 1, …,5, and when the model is not trained, the initial value of each element V is a random value within [0,1 ]. The network input F obtained by 5 neurons in the output layer uses the following algorithm:
F＝[f1,f2,f3,f4,f5]＝KV
elements Y of a network output vector Y of 5 neurons in a competition layeriThe calculation was performed using the following formula:
Figure BDA0002423308670000091
that is, in the output layer, the neuron which obtains the maximum input is in the excited state, the output value is 1, and the other neurons are in the inhibited state, and the output value is 0.
In the training process, the correction strategy from the input layer to the competition layer to connect the weight W is as follows:
(1) calculating a new connection weight W' according to the current connection weight W and the input vector:
W′＝W+α(X-W)
in the formula, 0< α <1 is learning efficiency.
(2) Normalizing W ', namely normalizing the element W'i,j(i-1, …, 11; j-1, …, Q) has:
Figure BDA0002423308670000092
(3) let W equal to W', the correction of W is achieved.
In the training process, the correction strategy of the connecting weight V from the competition layer to the output layer is as follows:
(1) calculating a new connection weight V' according to the current connection weight V and the input vector:
V′＝V+α(Y-V)
in the formula, 0< α <1 is learning efficiency.
(2) Normalizing V ', i.e. to element V'i,j(i-1, …, Q; j-1, …,5) has:
Figure BDA0002423308670000093
(3) let V equal V', the correction of V is achieved.
After the above training, the process proceeds to step S106, and it is determined whether the CPN network model converges. If the CPN network model does not converge, go to step S107, otherwise go to step S108.
And step S107 is entered, network parameters including the number Q of neurons in the competition layer, the learning efficiency alpha and the like are modified, and then step S105 is entered to retrain the CPN network model.
And S108, testing the trained CPN network model by using the test sample subset.
Then, step 109 is performed, whether the CPN network model accuracy meets the requirement is determined, if not, step S107 is performed, otherwise, step S110 is performed.
And step S110, storing the trained CPN network model.
Then, the process proceeds to step S111, where user satisfaction is evaluated based on the actual index variable observed value.
Fig. 2 shows a flowchart of a library user satisfaction comprehensive assessment calculation method based on a CPN network model according to an embodiment of the present invention. As shown in fig. 2, the method starts in step S201, where the evaluation index variable value is input, and the actual observed value of each index variable is input to the system to form an input vector X.
Then, the process proceeds to step S202, and reads the CPN network model. And reading the trained CPN network model parameters subjected to precision testing.
Subsequently, the process proceeds to step S203, and the CPN network model output is calculated. And calculating the output vector Y of the CPN network model according to the input vector X, the connection weight W from the input layer to the competition layer and the connection weight V from the competition layer to the output layer.
Next, the process proceeds to step S204, and the comprehensive evaluation result is displayed. And formatting and outputting a comprehensive evaluation result according to the output vector Y, wherein the corresponding relation between the output vector Y and the user satisfaction is as follows:
serial number Y Degree of satisfaction of user
1 [1,0,0,0,0] Height of
2 [0,1,0,0,0] Is higher than
3 [0,0,1,0,0] In general
4 [0,0,0,1,0] Is lower than
5 [0,0,0,0,1] Is low in
Further, the process proceeds to step S205, and it is inquired whether or not the result is saved. If the save result is selected, go to step S206, otherwise, directly exit the process.
Step S206 is entered, and the comprehensive evaluation result including the input vector X and the output vector Y is saved.
Then, the process proceeds to step S207, and it is queried whether to update the CPN network. If update is selected, go to step S208, otherwise, directly exit the process.
And S208, storing the current input vector X and the current output vector Y as new samples, incorporating the new samples into a system sample set to serve as training subset samples or testing subset samples, and retraining and testing the CPN network model. And finally, exiting the process.
The above description is only an embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes performed by the present specification and drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.

Claims (6)

1. A library user satisfaction comprehensive evaluation method based on a CPN network model is suitable for being executed in computer equipment, and is characterized in that: the method comprises the following steps:
setting user satisfaction index variables, and setting a plurality of user satisfaction index variables according to the evaluation standard of the user on the evaluation of the satisfaction of the user on the library;
setting a user satisfaction degree grade, wherein the user satisfaction degree grade comprises a plurality of grades from low to high;
constructing a CPN network model, taking the number of user satisfaction index variables as the number N of input layer neurons, taking the number of user satisfaction grades as the number M of output layer neurons, selecting the number Q of competitive layer neurons according to the number N of the input layer neurons and the number M of the output layer neurons, and constructing the CPN network model;
acquiring a sample set consisting of a user satisfaction index variable observed value and a user satisfaction value, training a CPN (compact peripheral component network) model by using the sample set, and evaluating the user satisfaction of the library by using the trained CPN model;
the number of competitive layer neurons Q is selected from a to-be-selected set U, and the to-be-selected set U is determined by the following formula:
Figure FDA0003579217330000011
wherein R (·) is a rounding function;
in the CPN network model, the network inputs obtained by Q neurons in the competition layer are:
Figure FDA0003579217330000012
wherein i is 1, …,11, j is 1, …, Q, X is an input vector, and W is a connection weight from the input layer to the contention layer; when the model is not trained, the initial value of each element of W is a random value in [0,1 ];
the network outputs of the Q neurons in the competition layer are: k ═ K1,…,kQ]；
In the formula, the ith neuron outputs a value kiCalculated as follows, i ═ 1, … …, Q:
Figure FDA0003579217330000021
in the output layer, the network input F obtained by M neurons adopts the following algorithm:
Figure FDA0003579217330000022
wherein i is 1, …, Q, j is 1, …,5, and V is the connection weight from the contention layer to the output layer; when the model is not trained, the initial value of each element of V is [0,1]]A random value of; elements Y of a network output vector Y of M neurons in the output layeriThe calculation was performed using the following formula:
Figure FDA0003579217330000023
the sample set comprises a training sample subset and a testing sample subset, wherein the training sample subset is used for training the CPN network model, the testing sample subset is used for testing the CPN network model, and the sample capacity of the training sample subset is larger than that of the testing sample subset; when the CPN model is applied to evaluate the satisfaction degree of the library user, the actual user satisfaction degree index variable is used as the input value of the CPN model to be calculated, the output value of the CPN model representing the comprehensive evaluation value of the user satisfaction degree is obtained, the comprehensive evaluation of the satisfaction degree of the library user is realized, a sample formed by the current index observation value and the comprehensive evaluation value of the user satisfaction degree is brought into a training sample subset or a testing sample subset, and the CPN model is updated at the same time.
2. The comprehensive evaluation method for the satisfaction degree of the library users based on the CPN network model as claimed in claim 1, wherein: the user satisfaction index variables comprise a plurality of primary index variables, and a plurality of secondary index variables are arranged under each primary index variable; the first-level index variables comprise three variables of facility environment conditions, service quality and resource construction; wherein the facility environmental condition comprises service system stability x1Service system response speed x2Network security and smoothness x3And resource acquisition facility x4Four secondary index variables; the quality of service comprises service flow normalization x5Service staff specialty x6And service content personalization x7Three secondary index variables; the resource construction comprises information resource authority x8Information content timeliness x9Resource structure completeness x10And information push accuracy x11And four secondary index variables, wherein the total number of the secondary index variables is used as the neuron number N of the input layer, and the input vector of the input layer is generated through each secondary index variable.
3. A library user satisfaction degree comprehensive assessment method based on CPN network model according to claim 2, characterized in that: the user satisfaction grades comprise five grades of low, lower, general, higher and higher from low to high in sequence.
4. The comprehensive evaluation method for the satisfaction degree of the library users based on the CPN network model as claimed in claim 1, wherein: when the input vector is transferred to the input layer of the CPN network model, the calculation value on the input layer is processed
Figure FDA0003579217330000031
And (5) performing normalization processing, wherein i is 1, … … and N.
5. The comprehensive evaluation method for the satisfaction degree of the library users based on the CPN network model as claimed in claim 1, wherein when the CPN network model is trained by adopting the subset of training samples, the correction strategy of the input layer to competition layer connection weight W is as follows:
calculating a new connection weight W' according to the current connection weight W and the input vector:
W′＝W+α(X-W)；
wherein 0< α <1 is learning efficiency;
w 'is normalized, and W is set to W', thereby correcting W.
6. The comprehensive evaluation method for the satisfaction degree of the library users based on the CPN network model according to claim 1, characterized in that when the CPN network model is trained by adopting the training sample subset, the correction strategy from the competition layer to the output layer connection weight V is as follows:
calculating a new connection weight V' according to the current connection weight V and the input vector:
V′＝V+α(Y-V)
wherein 0< α <1 is learning efficiency;
and carrying out normalization processing on V ', and enabling V to be equal to V', so as to realize the correction of V.
CN202010212527.6A 2020-03-24 2020-03-24 Comprehensive evaluation method for satisfaction degree of library users based on CPN network model Active CN111401764B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010212527.6A CN111401764B (en) 2020-03-24 2020-03-24 Comprehensive evaluation method for satisfaction degree of library users based on CPN network model

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010212527.6A CN111401764B (en) 2020-03-24 2020-03-24 Comprehensive evaluation method for satisfaction degree of library users based on CPN network model

Publications (2)

Publication Number Publication Date
CN111401764A CN111401764A (en) 2020-07-10
CN111401764B true CN111401764B (en) 2022-05-10

Family

ID=71431163

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010212527.6A Active CN111401764B (en) 2020-03-24 2020-03-24 Comprehensive evaluation method for satisfaction degree of library users based on CPN network model

Country Status (1)

Country Link
CN (1) CN111401764B (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2003344270A (en) * 2002-05-27 2003-12-03 Nippon Denro Kk System for evaluating deterioration of steel surface using self structurizing character map
CN102609764A (en) * 2012-02-01 2012-07-25 上海电力学院 CPN neural network-based fault diagnosis method for stream-turbine generator set
CN106447203A (en) * 2013-04-08 2017-02-22 江苏理工学院 Energy-saving assessment method for power supply and distribution network of industrial enterprise

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2003344270A (en) * 2002-05-27 2003-12-03 Nippon Denro Kk System for evaluating deterioration of steel surface using self structurizing character map
CN102609764A (en) * 2012-02-01 2012-07-25 上海电力学院 CPN neural network-based fault diagnosis method for stream-turbine generator set
CN106447203A (en) * 2013-04-08 2017-02-22 江苏理工学院 Energy-saving assessment method for power supply and distribution network of industrial enterprise

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
区域水资源可持续利用系统评价的模糊对向传播神经网络模型;程瑶等;《水文》;20080225(第01期);全文 *
基于神经网络的图书馆服务用户满意度评价体系;李雪萍等;《情报杂志》;20060318(第03期);全文 *
基于粗糙集-CPN网络的客户价值预测;梁娜等;《统计与决策》;20080310(第05期);全文 *
新建本科院校图书馆绩效评价研究;肖三霞;《价值工程》;20191231;全文 *
高校图书馆服务用户满意度的BP神经网络模型的建立;鞠建伟等;《情报杂志》;20040818(第08期);全文 *

Also Published As

Publication number Publication date
CN111401764A (en) 2020-07-10

Similar Documents

Publication Publication Date Title
CN109064023B (en) Method and device of manpower efficiency management system
CN106845777A (en) Case workload evaluation method, judge&#39;s performance evaluation techniques, modeling method, computing device and computer-readable recording medium
CN105956798A (en) Sparse random forest-based method for assessing running state of distribution network device
CN110599040A (en) Maintenance training evaluation method and system and terminal equipment
CN112861436A (en) Real-time prediction method for engine emission
CN114971227A (en) Power distribution network equipment risk assessment method based on MARCOS method
CN111209316A (en) Information literacy data mining method and device, electronic equipment and storage medium
CN113221332A (en) Coastal erosion vulnerability assessment method based on cloud model theory
CN112818525A (en) Hydroelectric generating set state evaluation method and system
CN117473424A (en) Transformer fault diagnosis method, system, equipment and medium based on random forest
CN115409317A (en) Transformer area line loss detection method and device based on feature selection and machine learning
CN113722898B (en) Durability evaluation method and device for fuel cell stack
CN107300907A (en) With reference to the flight control system Reliable Evaluating Methods of Their Performance of comprehensive assessment and hypothesis testing
CN111401764B (en) Comprehensive evaluation method for satisfaction degree of library users based on CPN network model
CN107644285A (en) The screening of power sales profitability evaluation index and Weight Determination and system
CN111680740A (en) Neural network training method and device and electrical load distinguishing method and device
CN116911635A (en) Low-voltage distribution network station running state evaluation method, device, terminal and medium
CN109657967A (en) A kind of confirmation method and system of Transmission Expansion Planning in Electric evaluating indexesto scheme weight
CN108596430A (en) A kind of professional Design of Shopping Website evaluation method
CN114091908A (en) Power distribution network comprehensive evaluation method, device and equipment considering multi-mode energy storage station
CN107679753A (en) A kind of electricity market principle market power evaluation method based on weighted Mahalanobis distance method
CN113537759A (en) User experience measurement model based on weight self-adaptation
CN112699920A (en) Method and system for determining main influence factors of line loss rate of passive station area
CN112348066A (en) Line uninterrupted power rating evaluation method based on gray clustering algorithm
CN106997472A (en) User characteristics sorting technique, user credit appraisal procedure and the device of user credit model

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