CN113064965A - Intelligent recommendation method for similar cases of civil aviation unplanned events based on deep learning - Google Patents
Intelligent recommendation method for similar cases of civil aviation unplanned events based on deep learning Download PDFInfo
- Publication number
- CN113064965A CN113064965A CN202110307398.3A CN202110307398A CN113064965A CN 113064965 A CN113064965 A CN 113064965A CN 202110307398 A CN202110307398 A CN 202110307398A CN 113064965 A CN113064965 A CN 113064965A
- Authority
- CN
- China
- Prior art keywords
- event
- vector
- word
- deep learning
- similarity
- 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
- 238000000034 method Methods 0.000 title claims abstract description 23
- 238000013135 deep learning Methods 0.000 title claims abstract description 19
- 238000012549 training Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000007781 pre-processing Methods 0.000 claims description 7
- 230000011218 segmentation Effects 0.000 claims description 7
- 238000013480 data collection Methods 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- 238000013507 mapping Methods 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 230000004044 response Effects 0.000 abstract description 7
- 238000004458 analytical method Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000012423 maintenance Methods 0.000 description 2
- 238000003058 natural language processing Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000007246 mechanism Effects 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
- 230000002035 prolonged effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/3332—Query translation
- G06F16/3335—Syntactic pre-processing, e.g. stopword elimination, stemming
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3346—Query execution using probabilistic model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Machine Translation (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a deep learning-based intelligent recommendation method for similar cases of civil aviation unplanned events, which is used for obtaining the similarity between different cases and a target event based on a Doc2Vec deep learning algorithm, quickly positioning the optimal case and providing a reference for formulating emergency treatment measures of the target event. The deep learning-based intelligent recommendation method for similar cases of civil aviation unplanned events can accurately and quickly perform similar case matching on target events, and realizes quick emergency response of unplanned events of a civil aircraft system.
Description
Technical Field
The invention belongs to the technical field of civil aviation safety, and particularly relates to an intelligent recommendation method for similar cases of an unplanned event of a civil aviation.
Background
Civil aviation safety is a permanent theme of the civil aviation industry, the civil aviation industry rises rapidly, and an airplane system is widely applied. Aviation maintenance decisions and airworthiness management also become key concerns for various civil aviation organizations. The quantity of civil aircraft equipment of each airline increases rapidly, the system complexity improves, and the factor that influences aircraft operation safety also increases by a wide margin. During the actual operation of an aircraft, various unplanned events are often accompanied. Although most of them are related to conventional problems, emergency response to unplanned events in aircraft operations is very necessary in order to avoid the development of unsafe conditions into serious safety problems caused by unplanned events. The emergency response of the unplanned event can timely and accurately correct the high-risk unsafe state, and the minimum casualties and economic losses are avoided.
Deep learning is a machine learning method simulating a human brain neural network mechanism, and achieves better effects in the image and vision fields. With the popularization and application of deep learning methods in Natural Language Processing (NLP), the solution of text analysis by deep learning algorithms has been widely studied. The deep learning method can dig out deep meanings of texts and improve the accuracy of extracting the key information of the texts. In the field of aviation safety, a deep learning algorithm is mature and applied to the aspects of potential safety hazards in the flight stage of an airplane, analysis of association relation between the potential safety hazards and accident reasons, analysis of major aviation accident reasons, analysis of reasons of aviation safety reports and the like. The research on intelligent recommendation of similar cases of civil aircraft unplanned events is less.
Disclosure of Invention
In order to solve the technical problems mentioned in the background technology, the invention provides an intelligent case recommendation method for the civil aviation unplanned events based on deep learning, so that the rapid emergency response of the civil aviation unplanned events is realized, and the continuous airworthiness time of the civil aviation is prolonged.
In order to achieve the technical purpose, the technical scheme of the invention is as follows:
the method for intelligently recommending the similar cases of the civil aviation unplanned events based on deep learning comprises the following steps:
(1) data collection: determining the starting time and the ending time of data collection, and collecting M fault event log reports about an aircraft engine in total of X airplanes with the model of Y of an airline company as sample data;
(2) data normalization: standardizing aviation terms involved in fault event log reports using a civil aviation dictionary; event information containing specific execution codes of an airline company is removed, and only M' pieces of standardized event information are reserved;
(3) preprocessing the standardized data;
(4) training the event text into a multi-dimensional vector by using a Doc2Vec model;
(5) calculating the similarity between the target event and the M' pieces of standardized event texts;
(6) and (4) taking the target event as a new case, putting the new case into a case library for similar case retrieval, and screening out a reference case similar to the new case.
Further, in step (3), the data preprocessing method is as follows:
(3a) event text information is arranged to form event standardized Chinese text corpora;
(3b) removing special characters which can influence the calculation speed and the test precision from the corpus;
(3c) removing irrelevant stop words which can influence the calculation of the text similarity in the corpus;
(3d) and performing word segmentation processing on the text by adopting a word segmentation tool.
Further, the specific process of step (4) is as follows:
(4a) training by adopting a PV-DM model, and randomly initializing two matrixes W and D of NxV dimension, wherein V represents the size of word traffic, N is the dimension of a set finally-desired word vector, and N is far smaller than V;
(4b) mapping each paragraph of the input corpus into a vector as a column vector of a matrix D; each word is mapped into a vector as a column vector of the matrix W;
(4c) for a piece of event information, a given sequence of words (w)j-k,wj-k+1,···wj+k) Total of C in the context1The word vector corresponding to each word is { wj-k,wj-k+1,···wj+kK is the window size, where C1K, word vector wj-k,wj-k+1,···wj+kThe dimension of is V1, for a given word wjDefine the paragraph vector of the sentence in which it is located as
(4d) Word vector of C words to be input corpus wj-k,wj-k+1,···wj+kThe sum is added with a paragraph vector representing the sentenceThen carrying out projection transformation and mapping to an N-dimensional combination vector h;
(4e) weighting and summing each combination vector h in W to obtain uj:
uj=γj'h
Wherein, γj' is a weight used in weighted summation, and the dimension is V multiplied by 1, j is 1, 2. V;
wherein U, b are parameters of a Softmax function;
(4f) u to be obtainedjAndprediction word w substituted into Softmax functionjProbability of occurrence:
(4g) obtaining an optimized objective function y by adopting a random gradient descent algorithm:
(4h) and (5) repeating the steps (4a) - (4g) to find the optimal word vector dimension N to obtain the trained Doc2Vec model.
Further, in step (5), cosine similarity is used to measure the similarity between each piece of event information in the event log report and all events before and after the event, and the similarity between the event information and other events is summed and averaged to obtain the similarity between the event and other events.
Further, the formula of the cosine similarity is as follows:
wherein sim (a, b) represents the similarity between two events,a vectorized representation of the representation event a,representing a vectorized representation of event b.
Adopt the beneficial effect that above-mentioned technical scheme brought:
the invention utilizes the characteristic recognition characteristic of deep learning to realize the rapid response of the unplanned event, assists in formulating the emergency disposal scheme of the unplanned event of the airplane system, and realizes the rapid positioning of the unplanned event of the airplane and the accurate elimination of the unplanned factors from the perspective of text semantic mining. The invention improves the emergency disposal efficiency of the civil aviation unplanned event and reduces the possibility of generating potential safety hazard by the airplane.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a diagram of a Doc2Vec model (PV-DM model);
FIG. 3 is a diagram of a Doc2Vec model (PV-DBOW model);
FIG. 4 is a diagram illustrating the best word vector dimensions for Doc2Vec model training in an embodiment.
Detailed Description
The technical scheme of the invention is explained in detail in the following with the accompanying drawings.
The diversity, complexity and subjectivity of manual judgment of the unplanned event increase the emergency response time of the unplanned event, and the untimely emergency response may cause serious safety problems such as casualties, airplane damage and the like. In order to solve the problem of emergency disposal efficiency of the civil aviation unplanned event, the invention designs an intelligent recommendation method for similar cases of the civil aviation unplanned event based on deep learning, which comprises the following steps as shown in figure 1:
step 1: data collection: determining the starting time and the ending time of data collection, and collecting M fault event log reports about an aircraft engine in total of X airplanes with the model of Y of an airline company as sample data;
step 2: data normalization: standardizing aviation terms involved in fault event log reports using a civil aviation dictionary; event information containing specific execution codes of an airline company is removed, and only M' pieces of standardized event information are reserved;
and step 3: preprocessing the standardized data;
and 4, step 4: training the event text into a multi-dimensional vector by using a Doc2Vec model;
and 5: calculating the similarity between the target event and the M' pieces of standardized event texts;
step 6: and (4) taking the target event as a new case, putting the new case into a case library for similar case retrieval, and screening out a reference case similar to the new case.
In this embodiment, specifically, in step 1, firstly, an aircraft with 10 models of a certain airline company as boeing 737 is collected through a civil aircraft maintenance operation control system and an airline mobile end platform in a manual entry manner, 142 fault event log reports about an aircraft engine from 10 months 2015 to 10 months 2020 are used as sample data, wherein each fault log relates to 18 attributes, of the 18 attributes, part of the attributes have no influence on a text similarity calculation result, and the attributes belong to unrelated attributes and are manually removed. And finally, selecting two attributes of ATA and event description as indexes for case retrieval from each piece of event information.
In the embodiment, specifically, in step 2, the aviation terms in the event description are standardized through the civil aviation dictionary, meanwhile, the event information containing the specific execution code of the airline company is manually rejected, and finally, only 134 standardized fault event logs are reserved.
In this embodiment, preferably, the data preprocessing method in step 3 is as follows:
(3a) the method comprises the following steps that (1) aviation company fault log information is formed into event standardized Chinese text corpora for experiments;
(3b) removing special characters which can influence the calculation speed and the test precision in the corpus, wherein the special characters comprise brackets, slashes and various punctuations;
(3c) the method for removing irrelevant stop words which influence the text similarity calculation in the corpus by using the Baidu stop vocabulary comprises the following steps: not only, but also, etc.;
(3d) and performing word segmentation on the text by adopting a jieba word segmentation tool.
Specifically, event description items in the fault event log are extracted, and special characters, stop words and word segmentation are removed. After the data preprocessing is finished, the processing result of each event is as follows: the 'unit reports that APU start is unsuccessful, FAULIT lamp is on' -the 'unit reports that APU start is unsuccessful, FAULIT lamp is on', and the sentence is divided into single words.
In this embodiment, preferably, the specific process of step 4 is as follows:
(4a) training by adopting a PV-DM model, and randomly initializing two matrixes W and D of NxV dimension, wherein V represents the size of word traffic, N is the dimension of a set finally-desired word vector, and N is far smaller than V;
(4b) mapping each paragraph of the input corpus into a vector as a column vector of a matrix D; each word is mapped into a vector as a column vector of the matrix W;
(4c) for a piece of event information, a given sequence of words (w)j-k,wj-k+1,···wj+k) Total of C in the context1The word vector corresponding to each word is { wj-k,wj-k+1,···wj+kK is the window size, where C1K, word vector wj-k,wj-k+1,···wj+kThe dimension of is V1, for a given word wjDefine the paragraph vector of the sentence in which it is located as
(4d) Word vector of C words to be input corpus wj-k,wj-k+1,···wj+kThe sum is added with a paragraph vector representing the sentenceThen carrying out projection transformation and mapping to an N-dimensional combination vector h;
(4e) weighting and summing each combination vector h in W to obtain uj:
uj=γj'h
Wherein, γj' is a weight used in weighted summation, and the dimension is V multiplied by 1, j is 1, 2. V;
wherein U, b are parameters of a Softmax function;
(4f) u to be obtainedjAndprediction word w substituted into Softmax functionjProbability of occurrence:
(4g) obtaining an optimized objective function y by adopting a random gradient descent algorithm:
(4h) and (5) repeating the steps (4a) - (4g) to find the optimal word vector dimension N to obtain the trained Doc2Vec model.
Specifically, the Doc2Vec model is shown in fig. 2(PV-DM model) and fig. 3(PV-DBOW model). In this embodiment, a genesis toolkit of python is used, the PV-DM model is adopted in the default Doc2Vec model algorithm, the optimal word vector dimension size is set to 400, as shown in FIG. 4, the window length Windows is set to 3, the minimum occurrence number of training words min-count is set to 2, the thread number Workers is set to 4, and the iteration number epoch in training is set to 15. On an Intel (R) core (TM) i7 processor, a well-trained Doc2Vec model is finally obtained.
In this embodiment, preferably, in step 5, cosine similarity is used to measure the similarity between each piece of event information in the event log report and all events before and after the event, and the similarity between the event information and other events is calculated by summing and averaging. The formula of the cosine similarity is as follows:
wherein sim (a, b) represents the similarity between two events,a vectorized representation of the representation event a,representing a vectorized representation of event b.
In this embodiment, specifically, the target event is put into a case library as a new case for similar case retrieval, and a similar case with similarity degree ranking 10 before the new case is selected, and the result is shown in table 1.
TABLE 1 Intelligent recommendation results for similar cases
The embodiments are only for illustrating the technical idea of the present invention, and the technical idea of the present invention is not limited thereto, and any modifications made on the basis of the technical scheme according to the technical idea of the present invention fall within the scope of the present invention.
Claims (5)
1. The method for intelligently recommending the similar cases of the civil aviation unplanned events based on deep learning is characterized by comprising the following steps of:
(1) data collection: determining the starting time and the ending time of data collection, and collecting M fault event log reports about an aircraft engine in total of X airplanes with the model of Y of an airline company as sample data;
(2) data normalization: standardizing aviation terms involved in fault event log reports using a civil aviation dictionary; event information containing specific execution codes of an airline company is removed, and only M' pieces of standardized event information are reserved;
(3) preprocessing the standardized data;
(4) training the event text into a multi-dimensional vector by using a Doc2Vec model;
(5) calculating the similarity between the target event and the M' pieces of standardized event texts;
(6) and (4) taking the target event as a new case, putting the new case into a case library for similar case retrieval, and screening out a reference case similar to the new case.
2. The deep learning-based intelligent recommendation method for similar cases of civil aviation unplanned events according to claim 1, characterized in that in the step (3), the data preprocessing method is as follows:
(3a) event text information is arranged to form event standardized Chinese text corpora;
(3b) removing special characters which can influence the calculation speed and the test precision from the corpus;
(3c) removing irrelevant stop words which can influence the calculation of the text similarity in the corpus;
(3d) and performing word segmentation processing on the text by adopting a word segmentation tool.
3. The deep learning-based intelligent recommendation method for similar cases of civil aviation unplanned events according to claim 1, characterized in that the specific process of the step (4) is as follows:
(4a) training by adopting a PV-DM model, and randomly initializing two matrixes W and D of NxV dimension, wherein V represents the size of word traffic, N is the dimension of a set finally-desired word vector, and N is far smaller than V;
(4b) mapping each paragraph of the input corpus into a vector as a column vector of a matrix D; each word is mapped into a vector as a column vector of the matrix W;
(4c) for a piece of event information, a given sequence of words (w)j-k,wj-k+1,…wj+k) Total of C in the context1The word vector corresponding to each word is { wj-k,wj-k+1,…wj+kK is the window size, where C1K, word vector wj-k,wj-k+1,…wj+kThe dimension of is V1, for a given word wjDefine the paragraph vector of the sentence in which it is located as
(4d) Word vector of C words to be input corpus wj-k,wj-k+1,…wj+kThe sum is added with a paragraph vector representing the sentenceThen carrying out projection transformation and mapping to an N-dimensional combination vector h;
(4e) weighting and summing each combination vector h in W to obtain uj:
uj=γj'h
Wherein, γj' is the weight used in weighted summation, and the dimension is V × 1, j is 1,2 … V;
wherein U, b are parameters of a Softmax function;
(4f) u to be obtainedjAndprediction word w substituted into Softmax functionjProbability of occurrence:
(4g) obtaining an optimized objective function y by adopting a random gradient descent algorithm:
(4h) and (5) repeating the steps (4a) - (4g) to find the optimal word vector dimension N to obtain the trained Doc2Vec model.
4. The intelligent deep learning-based case-like recommendation method for unplanned civil aviation events according to claim 1, wherein in step (5), cosine similarity is used to measure the similarity between each piece of event information in the event log report and all events before and after the event, and the similarity between the event information and other events is summed and averaged to obtain the similarity between the event and other events.
5. The deep learning-based intelligent recommendation method for similar cases of civil aviation unplanned events according to claim 4, wherein the cosine similarity is expressed by the following formula:
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110307398.3A CN113064965A (en) | 2021-03-23 | 2021-03-23 | Intelligent recommendation method for similar cases of civil aviation unplanned events based on deep learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110307398.3A CN113064965A (en) | 2021-03-23 | 2021-03-23 | Intelligent recommendation method for similar cases of civil aviation unplanned events based on deep learning |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113064965A true CN113064965A (en) | 2021-07-02 |
Family
ID=76563060
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110307398.3A Pending CN113064965A (en) | 2021-03-23 | 2021-03-23 | Intelligent recommendation method for similar cases of civil aviation unplanned events based on deep learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113064965A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114943227A (en) * | 2022-05-30 | 2022-08-26 | 中国银行股份有限公司 | Matching method and device of user story and test case |
CN117520484A (en) * | 2024-01-04 | 2024-02-06 | 中国电子科技集团公司第十五研究所 | Similar event retrieval method, system, equipment and medium based on big data semantics |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9037320B1 (en) * | 2014-01-29 | 2015-05-19 | The Boeing Company | Unscheduled maintenance disruption severity and flight decision system and method |
CN110263257A (en) * | 2019-06-24 | 2019-09-20 | 北京交通大学 | Multi-source heterogeneous data mixing recommended models based on deep learning |
CN110737837A (en) * | 2019-10-16 | 2020-01-31 | 河海大学 | Scientific research collaborator recommendation method based on multi-dimensional features under research gate platform |
-
2021
- 2021-03-23 CN CN202110307398.3A patent/CN113064965A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9037320B1 (en) * | 2014-01-29 | 2015-05-19 | The Boeing Company | Unscheduled maintenance disruption severity and flight decision system and method |
CN110263257A (en) * | 2019-06-24 | 2019-09-20 | 北京交通大学 | Multi-source heterogeneous data mixing recommended models based on deep learning |
CN110737837A (en) * | 2019-10-16 | 2020-01-31 | 河海大学 | Scientific research collaborator recommendation method based on multi-dimensional features under research gate platform |
Non-Patent Citations (2)
Title |
---|
申远;黄志良;胡彪;王适之;: "基于Doc2Vec和深度神经网络的战场态势智能推送研究", 《智能计算机与应用》, vol. 10, no. 01, 31 January 2020 (2020-01-31), pages 50 - 55 * |
申远;黄志良;胡彪;王适之;: "基于Doc2Vec和深度神经网络的战场态势智能推送研究", 智能计算机与应用, vol. 10, no. 01, pages 50 - 55 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114943227A (en) * | 2022-05-30 | 2022-08-26 | 中国银行股份有限公司 | Matching method and device of user story and test case |
CN117520484A (en) * | 2024-01-04 | 2024-02-06 | 中国电子科技集团公司第十五研究所 | Similar event retrieval method, system, equipment and medium based on big data semantics |
CN117520484B (en) * | 2024-01-04 | 2024-04-16 | 中国电子科技集团公司第十五研究所 | Similar event retrieval method, system, equipment and medium based on big data semantics |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108520343B (en) | Risk model training method, risk identification device, risk identification equipment and risk identification medium | |
CN107506389B (en) | Method and device for extracting job skill requirements | |
CN111966812B (en) | Automatic question answering method based on dynamic word vector and storage medium | |
CN112908436B (en) | Clinical test data structuring method, clinical test recommending method and device | |
CN113064965A (en) | Intelligent recommendation method for similar cases of civil aviation unplanned events based on deep learning | |
CN110727778A (en) | Intelligent question-answering system for tax affairs | |
CN101097570A (en) | Advertisement classification method capable of automatic recognizing classified advertisement type | |
CN111144068A (en) | Similar arbitration case recommendation method and device | |
CN111858842A (en) | Judicial case screening method based on LDA topic model | |
CN112182249A (en) | Automatic classification method and device for aviation safety report | |
CN110955767A (en) | Algorithm and device for generating intention candidate set list set in robot dialogue system | |
CN106294505A (en) | A kind of method and apparatus feeding back answer | |
CN116227584A (en) | Pre-training language model building method, system and medium for power industry | |
CN115659244A (en) | Fault prediction method, device and storage medium | |
CN111046177A (en) | Automatic arbitration case prejudging method and device | |
CN113722494A (en) | Equipment fault positioning method based on natural language understanding | |
CN114330103A (en) | Spacecraft anomaly relevance model training method based on-orbit anomaly information | |
CN112036179B (en) | Electric power plan information extraction method based on text classification and semantic frame | |
KR20200001936A (en) | Method for classifying safety document on construction site and Server for performing the same | |
CN110362828B (en) | Network information risk identification method and system | |
CN116629228A (en) | Standard element duplicate checking method based on text mining | |
CN111191029A (en) | AC construction method based on supervised learning and text classification | |
CN116340777A (en) | Training method of log classification model, log classification method and device | |
CN114328819A (en) | Power safety production hidden danger pre-control method based on knowledge graph | |
CN113656548A (en) | Text classification model interpretation method and system based on data envelope analysis |
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 |