CN113449884A - Intelligent operation and maintenance recommendation technology for performance equipment based on deep neural network - Google Patents
Intelligent operation and maintenance recommendation technology for performance equipment based on deep neural network Download PDFInfo
- Publication number
- CN113449884A CN113449884A CN202110726674.XA CN202110726674A CN113449884A CN 113449884 A CN113449884 A CN 113449884A CN 202110726674 A CN202110726674 A CN 202110726674A CN 113449884 A CN113449884 A CN 113449884A
- Authority
- CN
- China
- Prior art keywords
- equipment
- neural network
- deep neural
- maintenance
- recommendation
- 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.)
- Pending
Links
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 59
- 238000012423 maintenance Methods 0.000 title claims abstract description 54
- 238000005516 engineering process Methods 0.000 title claims abstract description 15
- 230000007246 mechanism Effects 0.000 claims abstract description 37
- 238000012549 training Methods 0.000 claims abstract description 23
- 238000013135 deep learning Methods 0.000 claims abstract description 12
- 238000003062 neural network model Methods 0.000 claims abstract description 8
- 238000007781 pre-processing Methods 0.000 claims abstract description 7
- 239000013598 vector Substances 0.000 claims description 18
- 239000011159 matrix material Substances 0.000 claims description 13
- 238000012360 testing method Methods 0.000 claims description 6
- 238000011160 research Methods 0.000 claims description 4
- 238000000034 method Methods 0.000 description 13
- 238000012545 processing Methods 0.000 description 4
- 241000282414 Homo sapiens Species 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 3
- 230000000007 visual effect Effects 0.000 description 3
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 2
- 238000013136 deep learning model Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
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
- G06Q10/20—Administration of product repair or maintenance
-
- 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
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Human Resources & Organizations (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Economics (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Strategic Management (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an intelligent operation and maintenance recommendation technology for performance equipment based on a deep neural network, which overcomes the problem that a recommendation algorithm based on deep learning in the prior art is lack of flexibility, and the trust of maintainers on recommendation results can be reduced in practical application, and comprises the following steps: constructing a data set, and performing characteristic preprocessing on data in the data set; constructing a deep neural network model, and setting each parameter of the deep neural network; sending the data into a deep neural network for training, randomly selecting certain equipment for maintenance, and recommending other equipment by a model; integrating an attention mechanism into the deep neural network, selecting the same equipment in the step S3, and recommending other equipment by the model; and increasing the weights of different data characteristics according to the actual situation, and calculating corresponding recommendation results. The features of the performance equipment are extracted by the deep neural network, and different weights are given to the features of the performance equipment by the attention mechanism, so that the system recommends equipment combinations.
Description
Technical Field
The invention relates to the technical field of intelligent recommendation algorithm design in the operation and maintenance process of stage performance equipment, in particular to a performance equipment intelligent operation and maintenance recommendation technology based on a deep neural network.
Background
With the rapid development of the cultural industry, the number of theaters is gradually increased, the requirement on performance equipment in the theaters is very high, the safety in the performance process can be ensured by the high-performance and high-precision equipment, and the stage effect of the performance can be enhanced. Therefore, the regular management and maintenance of performance equipment in theaters are very important. However, the number and kinds of performance equipment in a theater are large, and manual inspection and maintenance cause a great waste of time and energy. In addition, during manual maintenance, the price, the service life and other information of each type of equipment affect the priority selection of the maintenance personnel, the maintenance personnel are inundated with excessive information, and the maintenance personnel cannot accurately select the equipment to be maintained, which results in low efficiency of manual maintenance.
Therefore, when performing performance equipment maintenance, how to find useful information in the excessive information is a key research problem. The recommendation system has good performance in solving the information overload problem, and is defined as a decision strategy of a user in a complex information environment. The traditional recommendation methods are mainly divided into three types: content-based recommendation methods, collaborative filtering-based recommendation methods, and hybrid recommendation methods. Although the traditional recommendation method can obtain a good recommendation result to a certain extent, the traditional recommendation method cannot show good performance when the data is sparse. Conventional recommendation methods still face many difficulties and challenges, especially when dealing with complex auxiliary information.
In recent years, deep learning has become a useful approach in recommendation systems. The deep learning can effectively extract the basic characteristics of the data by mining and analyzing the depth of the data, so that the defects of the traditional recommendation method are made up to a certain extent, and higher-quality recommendation can be realized. Therefore, deep learning is also increasingly used in the recommendation field. For example, the invention named as an intelligent medical advice recommendation method and system based on deep learning is disclosed in the patent office 2020, 9 month and 22 days, and the publication number of the invention is CN110473636B, the recommendation method in the invention comprises the following steps: step (1): and constructing a patient condition information base. Step (2): setting a patient to be prescribed as a current patient, collecting the illness state information of the current patient, judging whether the current patient has the medical advice information, and if so, executing the step (4); otherwise, executing the step (3); and (3): determining a patient with the closest patient condition information in the patient condition information base, and recommending the medical advice of the closest patient as the initial medical advice of the current patient; and (4): and inputting the illness state information of the current patient into the trained deep learning model M, and outputting the deep learning model M as the recommended medical advice of the current patient. And (3) analyzing complex illness state information data of the patient by adopting a deep learning technology, and intelligently recommending appropriate medical advice. However, the recommendation algorithm based on deep learning lacks flexibility, which may reduce the confidence of the maintenance personnel on the recommendation result in practical application.
With the advent of attention mechanism networks, some scholars have incorporated attention mechanisms into deep learning recommendation systems. Attention mechanisms draw on the human visual mechanism and are widely used in image recognition, natural language processing, and other fields. Note that the mechanism can highlight more important features of all features and suppress the influence of other irrelevant factors by giving different weights to data features.
Disclosure of Invention
The invention aims to overcome the problem that a recommendation algorithm based on deep learning in the prior art is lack of flexibility, which can reduce the trust degree of a maintainer on a recommendation result in practical application, and provides an intelligent operation and maintenance recommendation technology of performance equipment based on a deep neural network.
In order to achieve the purpose, the invention adopts the following technical scheme: an intelligent operation and maintenance recommendation technology for performance equipment based on a deep neural network is characterized by comprising the following steps:
s1: constructing a data set about the situation of theater performance equipment, and performing characteristic preprocessing on data in the data set;
s2: constructing a deep neural network model, and setting each parameter of the deep neural network;
s3: sending the processed data in the step S1 into a deep neural network for training, randomly selecting a certain device for maintenance after the training is finished, and recommending other devices with similar characteristics to the device by the model, namely generating a first recommendation result;
s4: integrating an attention mechanism into the deep neural network, selecting the same equipment in the step S3 for maintenance, and recommending other equipment with similar characteristics to the equipment by the model, namely generating a second recommendation result;
s5: and increasing the weight of different data characteristics of the selected equipment to generate a corresponding recommendation result.
The method comprises the steps of firstly sorting and numbering the performance equipment of the theatre obtained by investigation and statistics, constructing a data set required by an experiment on the basis, and carrying out characteristic pretreatment on the data in the data set. And secondly, building a model, wherein the model is based on the deep neural network, and the deep neural network has the advantage that the deep neural network can well extract data characteristics. Attention is then integrated into the deep neural network to highlight more important of these features, making the recommendation more trustworthy. And sending the preprocessed data into the built model for training. When maintenance personnel randomly select one equipment, the model recommends other equipment similar to the equipment, the introduced attention mechanism can give different weights to equipment data characteristics, more important characteristics are highlighted, the influence of irrelevant factors is reduced, the recommendation result is more flexible and has reliability, the time and the energy required by manual maintenance are reduced, and the operation and maintenance efficiency of the equipment is improved.
Preferably, the step S1 is further expressed as:
s1.1: numbering and sorting theater performance equipment for research and statistics, and constructing a data set containing different characteristics of each equipment on the basis of the numbering and sorting;
s1.2: different characteristics are represented by category fields, and the data in the data set is subjected to characteristic preprocessing.
Each performance device is numbered, and subsequent data processing is facilitated. The influence degree on the characteristics is represented by a category field, and can be represented by numbers 0 to 5 in the invention, wherein 0 represents the minimum influence degree, and 5 represents the maximum influence degree, so that a data set required by an experiment is constructed on the basis of the influence degree.
Preferably, in step S1, the data set includes different characteristics of each device, including "price", "life" and "attention of maintenance personnel" of each device. Each performance device has different characteristics, such as "price", "lifetime", "attention of the maintenance personnel", etc., which may influence the selection of the maintenance personnel.
Preferably, the step S2 further includes:
s2.1: taking the category field as an index of a deep neural network embedding matrix, and respectively taking a vector representing the variety of the performance equipment and a vector representing the corresponding quantity of each type of performance equipment as the input of a deep neural network embedding layer;
s2.2: and setting the learning rate of deep learning, the number of fully-connected layers and the dimensionality of each layer.
The numbers 0 to 5 representing the categories are used as indexes of the deep neural network embedding matrix, the embedding layer can convert the index values into dense vectors with fixed sizes, and the vector codes can be updated, so that the input of the deep neural network is facilitated. And respectively processing the vectors representing the types of the performance equipment and the vectors representing the corresponding quantity of each type of performance equipment to be used as the input of the deep neural network embedding layer. In the invention, when the deep neural network is constructed, an Adam optimizer is preferably used for parameter optimization, Dropout layers are used for preventing overfitting, and then the learning rate of deep learning, the number of fully-connected layers and the dimension of each layer are set.
Preferably, the step S3 further includes:
s3.1: the processed data in the step S1 are sent to the deep neural network constructed in the step S2 for training, and a corresponding feature matrix is generated;
s3.2: randomly selecting a certain device as a device to be maintained, and calculating cosine similarity between a feature vector and a feature matrix of the device by using a deep neural network;
s3.3: the 5 devices in the feature matrix that are most similar to this feature vector are selected as the other devices having similar features to the device to be maintained, i.e. the first recommendation.
And calculating a first recommendation result by adopting the deep neural network, so that the deep neural network can effectively extract features. On the other hand, comparison with a second recommendation result after the attention mechanism is added is facilitated, and the attention mechanism can be proved to improve the reliability of the recommendation result.
Preferably, the step S4 further includes:
s4.1: integrating an attention mechanism into the constructed deep neural network, and sending the processed data in the step S1 into a model for training to obtain the loss of a training set and the loss of a testing set;
s4.2: selecting the same equipment as the equipment in the step S3 for maintenance, and calculating other equipment with similar characteristics with the equipment according to the same step in the step S3, namely a second recommendation result;
s4.3: and comparing the first recommendation result with the second recommendation result in the step, and judging whether the attention mechanism can improve the reliability of the recommendation result.
The attention mechanism draws on the human visual mechanism, and by giving different weights to data characteristics, more important information is highlighted, and the influence of irrelevant factors is suppressed.
Preferably, the step S5 is further expressed as:
selecting equipment needing maintenance, giving different weights to data characteristics of the equipment needing maintenance according to actual conditions, and recommending other equipment with characteristics similar to those of the equipment needing maintenance by adopting a deep neural network model integrated with an attention mechanism.
Selecting one device, calculating a corresponding recommendation result by the model, and obtaining another recommendation result after weighting the price characteristic of the device by using an attention mechanism; different recommendations are also obtained when this device "life" feature is weighted using an attention mechanism. The specific weight of which feature is used can be selected according to actual conditions.
Therefore, the invention has the following beneficial effects: 1. the intelligent recommendation system for the operation and the maintenance of the performance equipment is realized by adopting a deep neural network and an attention mechanism, the deep neural network and the attention mechanism are organically combined, the deep neural network is used for extracting and training the characteristics of the performance equipment, then the attention mechanism is used for endowing different weights to the characteristics of the performance equipment, and different weights are endowed to the data characteristics of the equipment, so that more important characteristics are highlighted, the influence of irrelevant factors is reduced, and the system recommends the equipment combination which possibly needs to be maintained by a maintainer; 2. when maintenance personnel select one of the devices for maintenance, the model recommends other devices similar to the device, and an attention mechanism is introduced to make the recommendation result more flexible and reliable, so that the time and energy required by manual maintenance are reduced, and the operation and maintenance efficiency of the device is improved.
Drawings
FIG. 1 is a flow chart of the operation of the present invention;
FIG. 2 is a graph of test loss for a training set;
FIG. 3 is a test loss plot for a test set.
Detailed Description
The invention is described in further detail below with reference to the following detailed description and accompanying drawings:
in the embodiment shown in fig. 1, an intelligent operation and maintenance recommendation technology for performing equipment based on a deep neural network can be seen, and the operation flow is as follows: step one, constructing a data set about the situation of theater performance equipment, and performing characteristic preprocessing on data in the data set; constructing a deep neural network model, and setting parameters of the deep neural network; step three, sending the processed data in the step one into a deep neural network for training, randomly selecting certain equipment for maintenance after the training is finished, and recommending other equipment with similar characteristics to the equipment by a model, namely generating a first recommendation result; step four, integrating an attention mechanism into the deep neural network, selecting the same equipment in the third step for maintenance, and recommending other equipment with similar characteristics to the equipment by the model, namely generating a second recommendation result; and fifthly, increasing the weight of different data characteristics of the selected equipment to generate a corresponding recommendation result.
The following further illustrates the technical solutions and technical effects of the present invention by specific examples, which are provided for explaining the present invention, but the present invention is not limited to the following examples, and is directed to an important feature that a deep neural network can effectively extract features and an attention mechanism can improve the reliability of a recommendation result.
The first step is as follows: constructing a data set about the situation of theater performance equipment, and performing characteristic preprocessing on data in the data set
Research and statistics are carried out on all performance equipment in a certain theater, each performance equipment is numbered, subsequent data processing is facilitated, and the results are as follows:
each performance device has different characteristics, such as "price", "lifetime", "attention of the maintenance personnel", etc., which may influence the selection of the maintenance personnel. The degree of influence on these features is represented by a category field, and in the present invention, the numbers 0 to 5 represent that the degree of influence is minimal, and 5 represents that the degree of influence is maximal, on the basis of which a data set required for an experiment is constructed.
The second step is that: constructing a deep neural network model and setting each parameter of the deep neural network
The numbers 0 to 5 representing the categories are used as indexes of the deep neural network embedding matrix, the embedding layer can convert the index values into dense vectors with fixed sizes, and the vector codes can be updated, so that the input of the deep neural network is facilitated. And respectively processing the vectors representing the types of the performance equipment and the vectors representing the corresponding quantity of each type of performance equipment to be used as the input of the deep neural network embedding layer. In the invention, when a deep neural network is constructed, an Adam optimizer is preferably used for parameter optimization, the learning rate is set to be 0.0005, Dropout layers are used for preventing overfitting, the number of fully-connected layers is set to be 3, and the dimension of each layer is half of that of the previous layer.
The third step: sending the processed data in the first step into a deep neural network for training, randomly selecting a certain device for maintenance after the training is finished, recommending other devices with similar characteristics to the device by a model, and generating a first recommendation result
And sending the preprocessed data into the deep neural network constructed in the second step for training to generate a corresponding feature matrix. Randomly selecting a certain device, for example, selecting the device with the number of 24, calculating the cosine similarity between the feature vector of the certain device and the feature matrix by the deep neural network, and selecting 5 devices with the most similar feature vector in the feature matrix as recommendation results, namely selecting the devices with the numbers of 23, 26, 73, 41 and 65. The model recommends the equipment with the numbers 23, 26, 73, 41 and 65 to the maintenance personnel, which is the first recommendation result obtained after only the deep neural network is used for training when a certain equipment is selected.
The fourth step: integrating the attention mechanism into a deep neural network, selecting the same equipment in the third step for maintenance, and recommending other equipment with similar characteristics to the equipment by the model, namely generating a second recommendation result
The attention mechanism is fused into the constructed deep neural network, the attention mechanism refers to the visual mechanism of human beings, and different weights are given to data characteristics, so that more important information is highlighted, and the influence of irrelevant factors is suppressed. After the attention mechanism is integrated into the deep neural network, the processed data is sent to the model again for training, and the loss of the training set and the loss of the testing set are respectively shown in fig. 2 and fig. 3. And selecting the equipment with the number of 24 again, and calculating other equipment with similar characteristics with the equipment according to the same step in the third step, namely the equipment with the recommendation numbers of 23, 17, 53, 41 and 57, wherein the equipment is a second recommendation result obtained by training the deep neural network integrated with the attention mechanism when the same equipment is selected.
The fifth step: increasing the weight of different data characteristics of the selected device to generate corresponding recommendation results
Selecting equipment needing maintenance, giving different weights to data characteristics of the equipment needing maintenance according to actual conditions, and recommending other equipment with characteristics similar to those of the equipment needing maintenance by adopting a deep neural network model integrated with an attention mechanism.
If the device with the number of 11 is selected, the corresponding devices, namely the devices with the model recommendation numbers of 88, 78, 49, 4 and 33 are recommended. When the device 'price' feature is weighted using the attention mechanism, the model recommendations are for devices with recommendation numbers 10, 78, 27, 87, 28. When this device "life" feature is weighted using the attention mechanism, the model recommendations are for devices numbered 87, 4, 33, 28, 10.
The above-described embodiments are only preferred embodiments of the present invention, and are not intended to limit the present invention in any way, and other variations and modifications may be made without departing from the spirit of the invention as set forth in the claims.
Claims (7)
1. An intelligent operation and maintenance recommendation technology for performance equipment based on a deep neural network is characterized by comprising the following steps:
s1: constructing a data set about the situation of theater performance equipment, and performing characteristic preprocessing on data in the data set;
s2: constructing a deep neural network model, and setting each parameter of the deep neural network;
s3: sending the processed data in the step S1 into a deep neural network for training, randomly selecting a certain device for maintenance after the training is finished, and recommending other devices with similar characteristics to the device by the model, namely generating a first recommendation result;
s4: integrating an attention mechanism into the deep neural network, selecting the same equipment in the step S3 for maintenance, and recommending other equipment with similar characteristics to the equipment by the model, namely generating a second recommendation result;
s5: and increasing the weight of different data characteristics of the selected equipment to generate a corresponding recommendation result.
2. The technology of claim 1, wherein the step S1 is further represented as:
s1.1: numbering and sorting theater performance equipment for research and statistics, and constructing a data set containing different characteristics of each equipment on the basis of the numbering and sorting;
s1.2: different characteristics are represented by category fields, and the data in the data set is subjected to characteristic preprocessing.
3. The technology for intelligent operation and maintenance recommendation of performance equipment based on deep neural network as claimed in claim 2, wherein in step S1.1, said different characteristics of each device include "price", "life" and "attention of maintenance personnel" of each device.
4. The technology of claim 2, wherein the step S2 further includes:
s2.1: taking the category field as an index of a deep neural network embedding matrix, and respectively taking a vector representing the variety of the performance equipment and a vector representing the corresponding quantity of each type of performance equipment as the input of a deep neural network embedding layer;
s2.2: and setting the learning rate of deep learning, the number of fully-connected layers and the dimensionality of each layer.
5. The technology of claim 1, wherein the step S3 further includes:
s3.1: the processed data in the step S1 are sent to the deep neural network constructed in the step S2 for training, and a corresponding feature matrix is generated;
s3.2: randomly selecting a certain device as a device to be maintained, and calculating cosine similarity between a feature vector and a feature matrix of the device by using a deep neural network;
s3.3: the 5 devices in the feature matrix that are most similar to this feature vector are selected as the other devices having similar features to the device to be maintained, i.e. the first recommendation.
6. The technology of claim 1, wherein the step S4 is further represented as:
s4.1: integrating an attention mechanism into the constructed deep neural network, and sending the processed data in the step S1 into a model for training to obtain the loss of a training set and the loss of a testing set;
s4.2: selecting the same equipment as the equipment in the step S3 for maintenance, and calculating other equipment with similar characteristics with the equipment according to the same step in the step S3, namely a second recommendation result;
s4.3: and comparing the first recommendation result with the second recommendation result in the step, and judging whether the attention mechanism can improve the reliability of the recommendation result.
7. The technology of claim 1, wherein the step S5 is further represented as:
selecting equipment needing maintenance, giving different weights to data characteristics of the equipment needing maintenance according to actual conditions, and recommending other equipment with characteristics similar to those of the equipment needing maintenance by adopting a deep neural network model integrated with an attention mechanism.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110726674.XA CN113449884A (en) | 2021-06-29 | 2021-06-29 | Intelligent operation and maintenance recommendation technology for performance equipment based on deep neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110726674.XA CN113449884A (en) | 2021-06-29 | 2021-06-29 | Intelligent operation and maintenance recommendation technology for performance equipment based on deep neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113449884A true CN113449884A (en) | 2021-09-28 |
Family
ID=77813849
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110726674.XA Pending CN113449884A (en) | 2021-06-29 | 2021-06-29 | Intelligent operation and maintenance recommendation technology for performance equipment based on deep neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113449884A (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112084407A (en) * | 2020-09-08 | 2020-12-15 | 辽宁工程技术大学 | Collaborative filtering recommendation method fusing graph neural network and attention mechanism |
| CN112256980A (en) * | 2020-10-23 | 2021-01-22 | 辽宁工程技术大学 | Dynamic graph attention network-based multi-relation collaborative filtering recommendation |
| CN112989064A (en) * | 2021-03-16 | 2021-06-18 | 重庆理工大学 | Recommendation method for aggregating knowledge graph neural network and self-adaptive attention |
-
2021
- 2021-06-29 CN CN202110726674.XA patent/CN113449884A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112084407A (en) * | 2020-09-08 | 2020-12-15 | 辽宁工程技术大学 | Collaborative filtering recommendation method fusing graph neural network and attention mechanism |
| CN112256980A (en) * | 2020-10-23 | 2021-01-22 | 辽宁工程技术大学 | Dynamic graph attention network-based multi-relation collaborative filtering recommendation |
| CN112989064A (en) * | 2021-03-16 | 2021-06-18 | 重庆理工大学 | Recommendation method for aggregating knowledge graph neural network and self-adaptive attention |
Non-Patent Citations (1)
| Title |
|---|
| HUIMIN LI: "An Intelligent Recommendation System for Performance Equipment Operation and Maintenance via Deep Neural Network and Attention Mechanism", 《2021 IEEE 10TH DATA DRIVEN CONTROL AND LEARNINGSYSTEMS CONFERENCE》, 25 June 2021 (2021-06-25), pages 1464 - 1468 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110298037B (en) | Convolutional neural network matching text recognition method based on enhanced attention mechanism | |
| CN110705301B (en) | Entity relationship extraction method and device, storage medium and electronic equipment | |
| CN110781298B (en) | Medicine classification method, apparatus, computer device and storage medium | |
| CN107491729B (en) | Handwritten digit recognition method based on cosine similarity activated convolutional neural network | |
| CN111125520B (en) | Event line extraction method based on deep clustering model for news text | |
| CN113947161A (en) | Attention mechanism-based multi-label text classification method and system | |
| CN113628059A (en) | Associated user identification method and device based on multilayer graph attention network | |
| CN112529638B (en) | Service demand dynamic prediction method and system based on user classification and deep learning | |
| CN110046264A (en) | A kind of automatic classification method towards mobile phone document | |
| CN112418320B (en) | Enterprise association relation identification method, device and storage medium | |
| CN113593661A (en) | Clinical term standardization method, device, electronic equipment and storage medium | |
| CN112347223A (en) | Document retrieval method, document retrieval equipment and computer-readable storage medium | |
| CN116561264A (en) | Knowledge graph-based intelligent question-answering system construction method | |
| CN117494051A (en) | Classification processing method, model training method and related device | |
| CN113705242B (en) | Intelligent semantic matching method and device for education consultation service | |
| CN113722494A (en) | Equipment fault positioning method based on natural language understanding | |
| CN113449884A (en) | Intelligent operation and maintenance recommendation technology for performance equipment based on deep neural network | |
| CN115618297A (en) | Method and device for identifying abnormal enterprise | |
| Brummerloh et al. | Boromir at Touché 2022: Combining Natural Language Processing and Machine Learning Techniques for Image Retrieval for Arguments. | |
| CN117237857B (en) | Video understanding task execution method and device, storage medium and electronic equipment | |
| CN115905533B (en) | Multi-label text intelligent classification method | |
| CN114328797B (en) | Content search method, device, electronic apparatus, storage medium, and program product | |
| CN117690178B (en) | Face image recognition method and system based on computer vision | |
| CN116595158B (en) | User interaction method and system based on transducer model | |
| CN113821642B (en) | Method and system for cleaning text based on GAN clustering |
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 |