CN113535978B - Construction method of rail transit system operation risk ontology - Google Patents
Construction method of rail transit system operation risk ontology Download PDFInfo
- Publication number
- CN113535978B CN113535978B CN202110791328.XA CN202110791328A CN113535978B CN 113535978 B CN113535978 B CN 113535978B CN 202110791328 A CN202110791328 A CN 202110791328A CN 113535978 B CN113535978 B CN 113535978B
- Authority
- CN
- China
- Prior art keywords
- risk
- accident
- word
- concept
- vocabulary
- 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
Links
- 238000010276 construction Methods 0.000 title claims abstract description 22
- 238000000034 method Methods 0.000 claims abstract description 32
- 230000011218 segmentation Effects 0.000 claims description 15
- 238000012545 processing Methods 0.000 claims description 11
- 238000004364 calculation method Methods 0.000 claims description 9
- 239000011229 interlayer Substances 0.000 claims description 6
- 238000005516 engineering process Methods 0.000 claims description 5
- 239000010410 layer Substances 0.000 claims description 5
- 230000000694 effects Effects 0.000 claims description 3
- 230000007613 environmental effect Effects 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000011160 research Methods 0.000 abstract description 10
- 230000002265 prevention Effects 0.000 abstract description 3
- 239000007787 solid Substances 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 10
- 238000000605 extraction Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 6
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
- G06F40/216—Parsing using statistical 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
- 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
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- 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
- G06Q50/40—Business processes related to the transportation industry
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- General Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- General Engineering & Computer Science (AREA)
- Educational Administration (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Game Theory and Decision Science (AREA)
- Animal Behavior & Ethology (AREA)
- Development Economics (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Artificial Intelligence (AREA)
- Probability & Statistics with Applications (AREA)
- Primary Health Care (AREA)
- Machine Translation (AREA)
Abstract
The invention provides a construction method of a rail transit system operation risk body. The method comprises the following steps: according to the operation accident characteristics of the rail transit system, starting from four aspects of a man-machine loop, constructing a five-class sub word library representing the accident characteristics, and setting weights of the five-class sub word library; calculating risk concept words of the accident data text by utilizing improved TF-IDF based on the accident data text and the five-class word library, calculating weight values of the risk concept words according to weights corresponding to the five-class word library, and extracting hierarchical and non-hierarchical relations among the risk concept words by utilizing an improved K-means algorithm; and constructing the rail transit system operation risk ontology according to the risk concepts of the rail transit operation risk ontology, the hierarchical and non-hierarchical relations among the risk concept words and the weight values of the risk concept words. The method can effectively extract the text risk of the urban rail transit operation accident, provides important data support for the rail transit system operation risk research, and lays a solid foundation for the risk prevention and control research.
Description
Technical Field
The invention relates to the technical field of rail transit system operation safety management, in particular to a construction method of a rail transit system operation risk body.
Background
With the continuous acceleration of urban construction in China, the increasing travel demands of people accelerate the development of urban rail transit. As a convenient, safe and comfortable urban public transport means, urban rail transit has a great number of advantages, but various potential safety hazards exist in the urban rail transit, and the urban rail transit also attracts wide attention.
In recent years, most cities in China have been provided with rail transportation, and urban rail transportation belongs to passenger-intensive public places, and most of the urban rail transportation runs in underground closed spaces, if a security event occurs, great social influence is generated, so that the security of urban rail transportation operation is ensured to be the primary working focus of operation enterprises. In view of the fact that urban rail transit projects of most cities belong to new construction or reconstruction and extension, operation safety management experience is relatively lacking, new technologies are applied, new personnel are added, local society is unstable, and networked operation brings new challenges to urban rail transit operation safety management work.
The urban rail transit security risk prediction is one of the cores of urban rail transit security management and control work, the historical events and data of urban rail transit operation security, and the security work experience of operation management companies, expert knowledge in the field is an important component part of urban rail transit operation security risk prediction, the urban rail transit operation is strong in specialization, the equipment technology is complex, and the number of passengers participates in the urban rail transit is large, so that the risk prediction difficulty is high, and the method is a typical uncertainty knowledge expression and reasoning process.
At present, an effective construction method for an operation risk body of a rail transit system is not available in the prior art.
Disclosure of Invention
The embodiment of the invention provides a construction method of a rail transit system operation risk body.
In order to achieve the above purpose, the present invention adopts the following technical scheme.
A construction method of a rail transit system operation risk ontology comprises the following steps:
according to the operation accident characteristics of the rail transit system, starting from four aspects of a man-machine loop, constructing a five-class sub word library representing the accident characteristics, and setting weights of the five-class sub word library;
calculating risk concept words of the accident data text by utilizing improved TF-IDF based on the accident data text of the rail transit operation and the five-class sub word library, calculating weight values of the risk concept words according to weights corresponding to the five-class sub word library, and extracting hierarchical and non-hierarchical relations among the risk concept words by utilizing an improved K-means algorithm;
and constructing the rail transit system operation risk ontology according to the risk concepts of the rail transit operation risk ontology, the hierarchical and non-hierarchical relations among the risk concept words and the weight values of the risk concept words.
Preferably, the constructing the five-class sub word library for representing accident characteristics according to the operation accident characteristics of the rail transit system from four aspects of man-machine loop comprises:
from the accident feature of the rail transit system operation, dividing the accident professional word stock of the urban rail transit operation into four sub word stocks of personnel, physical component class, environment class and safety management class;
respectively using formulas (1-1), (1-2), (1-3), (1-4) and (1-5) to represent vocabulary sets in the accident professional word stock, the personnel sub word stock, the physical group classification sub word stock, the environment sub word stock and the safety management sub word stock;
V＝{V P ,V S ,V E ,V M } (1-1)
V P ＝{V P1 ,V P2 ,V P3 ,...,V Pn } (1-2)
V S ＝{V S1 ,V S2 ,V S3 ,...,V Sm } (1-3)
V E ＝{V E1 ,V E2 ,V E3 ,...,V Ei } (1-4)
V M ＝{V M1 ,V M2 ,V M3 ,...,V Mj } (1-5)
v in P -a collection of staff-like vocabulary;
V S -a collection of physical group classification words;
V E -an environment-like vocabulary set;
V M -managing a collection of class vocabularies;
n-total number of personnel vocabulary;
m-total number of physical group classification vocabulary;
i-the total number of environmental vocabulary;
j-manage the total number of class vocabularies.
Preferably, the weight setting for the five-class sub word library includes:
and (3) calculating the proportion of the vocabulary in each sub word stock in the whole accident professional word stock by using formulas (1-6), (1-7), (1-8) and (1-9) respectively.
Assume that
Same reason
Same reason
Same reason
W(V PX ) Vocabulary V PX In vocabulary library V P The weight of (a);
S A (V PX ) -by vocabulary V PX The sub word bank V P The number of occurrences of the accident caused by the vocabulary in (a);
W(V SX ) Vocabulary V SX In vocabulary library V S The weight of (a);
S A (V SX ) -by vocabulary V SX The sub word bank V S The number of occurrences of the accident caused by the vocabulary in (a);
W(V EX ) Vocabulary V EX In vocabulary library V E The weight of (a);
S A (V EX ) -by vocabulary V EX The sub word bank V E The number of occurrences of the accident caused by the vocabulary in (a);
W(V MX ) Vocabulary V MX In vocabulary library V M The weight of (a);
S A (V MX ) -by vocabulary V MX The sub word bank V M The number of occurrences of the accident caused by the vocabulary in (a);
calculating the weight of any word in the sub-class word stock by using the formula (1-10);
v in Zx -any vocabulary in the lexicon V;
W(V Zx ) Vocabulary V Zx Weights in word stock V;
S A (V Zx ) -by vocabulary V Zx Number of incidents caused by vocabulary in the sub-word library
S A (V) -the vocabulary in each sub-class word library results in a total number of incidents.
Preferably, the accident data text based on the rail traffic operation and the five-class sub word library calculate a risk concept word of the accident data text by using an improved TF-IDF, and calculate a weight value of the risk concept word according to a weight corresponding to the five-class sub word library, including:
performing word segmentation processing on the accident data text by using a Chinese text word segmentation technology, and optimizing word segmentation effect of the accident data text in a mode of combining hidden Markov and Viterbi algorithm to obtain an accident word segmentation set;
and respectively traversing the accident professional word stock for each accident word, judging whether each accident word belongs to the four sub word stocks, if so, determining that each accident word is a risk concept word of the accident data text, comprehensively considering the weight of each sub word stock in the accident professional word stock and the weight of the vocabulary in each sub word stock in the whole accident professional word stock, calculating the weight value of the risk concept word of the accident data text by utilizing a TF-IDF algorithm, and sequencing the weight values of the risk concept words.
Preferably, the calculating the weight value of the risk concept word of the accident data text by TF-IDF algorithm comprehensively considering the weight of each sub word stock in the accident professional word stock and the weight of the vocabulary in each sub word stock in the whole accident professional word stock includes:
the calculation modes of the weight values of the risk concept words are formulas (1-11) and (1-12).
And, in addition, the method comprises the steps of,
δ+γ＝1 (1-12)
delta-coefficient factor for measuring comprehensive weight of risk feature words, wherein delta is more than 0 and less than 1;
gamma, a coefficient factor for measuring comprehensive weight of risk feature words, wherein gamma is more than 0 and less than 1.
Preferably, the extracting the hierarchical and non-hierarchical relationship between the risk concept words by using the improved K-means algorithm includes:
dividing the relationship among the risk concept words into a longitudinal hierarchical relationship reflecting the risk concept hierarchy and a transverse non-hierarchical relationship reflecting the risk concept association in the risk ontology model, wherein the longitudinal hierarchical relationship describes the upper and lower relationship and father-son relationship among the risk concepts and indicates the subordinate relationship among the risks, and the non-hierarchical relationship refers to other connection relationships among the risk concept words except the hierarchical relationship;
the top-down modeling principle is followed, the top-level concept words are selected by comprehensively considering the relation between the accident text set and the risk ontology concept words, and the following formula is used for selecting the basis for the risk ontology top-level concept words:
k is the total amount of accident text;
m-total number of risk ontology concept words;
C x ,C y -representing arbitrary conceptual words;
f(C xi ,C yi ) Concept word C x ，C y The frequency of occurrence is the same as the ith accident text;
F(C x ,C y ) Concept word C x ，C y Co-occurrence frequency in accident text set;
the calculation formula of the interlayer relation among the risk ontology concept words is as follows:
in the formula, con (C) x ,C y ) Concept word C x ，C y Confidence between concepts used for representing interlayer relations between concepts;
F(C x ) -express concept word C x Co-occurrence frequency in accident text set;
the processing procedure for extracting the hierarchical and non-hierarchical relations between the risk concept words by using the improved K-means algorithm comprises the following steps:
step1: determining top-level risk concept words in the risk concept word set according to formulas (1-14);
step2: establishing a confidence relation between the top-level risk concept words and other risk concept words according to the formula (1-15), and judging the number of sub-risk concept words of the top-level risk concept words;
step3: if the confidence level is greater than the threshold, i.e. Con (C t ,G x ) Judging that a connection relationship exists between the sub-risk concept word and the top-level risk concept word, and constructing a space vector between the sub-risk concept word and the top-level risk concept word;
step4: if the confidence level is less than the threshold, i.e. Con (C t ,G x ) Judging that no connection relation exists between the sub risk concept words and the top risk concept words if T is less than or equal to T;
step5: clustering by using an improved K-means algorithm to obtain a cluster with Z= { Z mi I=1, 2,3,..k } is a risk concept cluster of cluster centers, namely:
Cluster＝{Z 1 {C 1 ,C 2 ,...,C x },...,Z i {C 1 ,C 2 ,...,C y },...,Z k {C 1 ,C 2 ,...,C z }}；
step6: repeating the steps, and clustering each cluster until the cluster center is not changed any more;
step7: judging top-level risk concept words in each clustered risk concept cluster, and taking the top-level risk concept words as child node risk concept words of the top-level risk concept words of the upper layer;
step8: steps 3 through 7 are repeated until the determination of the relationships between all risk concept words is completed.
Preferably, the constructing the rail transit system operation risk ontology according to the risk concepts of the rail transit operation risk ontology, the hierarchical and non-hierarchical relationships between the risk concept words, and the weight values of the risk concept words includes:
according to the hierarchical and non-hierarchical relationship between risk concept words, the connection between related risk concepts with hierarchical relationship and non-hierarchical relationship with the risk concept words on the top layer of the risk ontology is completed, and the preliminary construction of the operation risk ontology structure of the rail transit system is realized;
the method comprises the steps of completing structural description of hierarchical and non-hierarchical relations among risk concept words by representing attributes of the risk concept words in a risk ontology, and constructing a hierarchical framework of an operation risk ontology of a rail transit system, wherein the attributes of the risk concept words comprise object attributes and data attributes;
and according to the initially constructed rail transit system operation risk body structure and the hierarchical framework of the rail transit system operation risk body, the construction of the rail transit system operation risk body is completed.
According to the technical scheme provided by the embodiment of the invention, the text risk of the urban rail transit operation accident can be effectively extracted, and meanwhile, in order to further utilize the risk related information in a limited way, the description provides a risk ontology model capable of representing the operation accident, and effective storage of risk knowledge is completed in a structured, formal and standardized form. The method provides important data support for the rail transit system operation risk research, and lays a solid foundation for the risk prevention and control research.
Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a process flow diagram of a bayesian reasoning method based on a rail transit system operation risk ontology provided by an embodiment of the present invention;
fig. 2 is a schematic diagram of a processing procedure of extracting risk concept words of accident data text according to an embodiment of the present invention;
fig. 3 is a schematic diagram of extracting top-level risk concept words and inter-word clusters according to an embodiment of the present invention.
FIG. 4 is a schematic diagram of a intra-cluster risk concept word screening, a top-level risk concept connection between each cluster and a risk concept word relationship structure provided by an embodiment of the present invention;
FIG. 5 is a block diagram of a top-level risk concept class according to an embodiment of the present invention;
fig. 6 is a schematic diagram of a part of an ontology description language for constructing a top-level risk concept class according to an embodiment of the present invention.
Detailed Description
Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals refer to the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary only for explaining the present invention and are not to be construed as limiting the present invention.
As used herein, the singular forms "a", "an", "the" and "the" are intended to include the plural forms as well, unless expressly stated otherwise, as understood by those skilled in the art. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element or intervening elements may also be present. Further, "connected" or "coupled" as used herein may include wirelessly connected or coupled. The term "and/or" as used herein includes any and all combinations of one or more of the associated listed items.
It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
For the purpose of facilitating an understanding of the embodiments of the invention, reference will now be made to the drawings of several specific embodiments illustrated in the drawings and in no way should be taken to limit the embodiments of the invention.
The embodiment of the invention discloses a Bayesian inference method based on a rail transit system operation risk ontology, which comprises the following steps: from four aspects of a man-machine loop, establishing a professional word stock for representing the operation accident characteristics of the system and setting weights; an operation risk knowledge extraction method utilizing a hidden Markov and Viterbi algorithm and an improved K-means algorithm is provided; and constructing the rail transit system operation risk ontology according to the risk concepts of the risk ontology, the risk concept relations and the attributes among the risk concepts.
The processing flow of the Bayesian inference method based on the rail transit system operation risk ontology provided by the embodiment of the invention is shown in the figure 1, and comprises the following processing steps:
and S10, constructing a five-class sub word library representing accident characteristics according to the operation accident of the rail transit system, and setting weights of the five-class sub word libraries of each class.
And S20, calculating risk concept words of the accident data text by utilizing the improved TF-IDF based on the accident data text operated by the rail transit and the five-class sub word library, and calculating weight values of the risk concept words according to weights corresponding to the five-class sub word library. And extracting the hierarchy and non-hierarchy relation among the risk concepts of the risk concept words by using an improved K-means algorithm.
And step S30, constructing a rail transit system operation risk ontology according to the risk concepts of the rail transit operation risk ontology, the hierarchical and non-hierarchical relations among the risk concepts of the risk concept words and the weight values of the risk concept words.
The step S10 specifically comprises the steps of constructing five-class sub word libraries from four aspects of a man-machine loop based on the operation accident characteristics of the rail transit system, respectively setting different weight coefficients for various word libraries according to the difference of the occurrence probability of any accident of the vocabulary in the various sub word libraries, and further perfecting the construction of the various word libraries.
The accident professional word stock of urban rail transit operation is divided into four sub word stock of personnel, physical component, environment and safety management. Because the urban rail transit operation accidents do not occur according to the probabilities of four reasons, namely a man-machine loop, the weights of the sub word banks in the accident professional word bank are different.
The vocabulary sets in the accident professional word stock, the personnel sub word stock, the physical group classification sub word stock, the environment sub word stock and the safety management sub word stock are respectively represented by formulas (1-1), (1-2), (1-3), (1-4) and (1-5).
V＝{V P ,V S ,V E ,V M } (1-1)
V P ＝{V P1 ,V P2 ,V P3 ,...,V Pn } (1-2)
V S ＝{V S1 ,V S2 ,V S3 ,...,V Sm } (1-3)
V E ＝{V E1 ,V E2 ,V E3 ,...,V Ei } (1-4)
V M ＝{V M1 ,V M2 ,V M3 ,...,V Mj } (1-5)
V in P -a collection of staff-like vocabulary;
V S -a collection of physical group classification words;
V E -an environment-like vocabulary set;
V M -managing a collection of class vocabularies;
n-total number of personnel vocabulary;
m-total number of physical group classification vocabulary;
i-the total number of environmental vocabulary;
j-manage the total number of class vocabularies.
Considering that the accident occurrence frequency is different due to the vocabulary of each sub-word bank, the proportion of the vocabulary in each sub-word bank in the whole accident professional word bank is calculated by using formulas (1-6), (1-7), (1-8) and (1-9) according to the different accident data proportion of different accident types.
Assume that
Same reason
Same reason
Same reason
W(V PX ) Vocabulary V PX In vocabulary library V P The weight of (a);
S A (V PX ) -by vocabulary V PX The sub word bank V P The number of occurrences of the accident caused by the vocabulary in (a);
W(V SX ) Vocabulary V SX In vocabulary library V S The weight of (a);
S A (V SX ) -by vocabulary V SX The sub word bank V S The number of occurrences of the accident caused by the vocabulary in (a);
.W(V EX ) Vocabulary V EX In vocabulary library V E The weight of (a);
S A (V EX ) -by vocabulary V EX The sub word bank V E The number of occurrences of the accident caused by the vocabulary in (a);
W(V MX ) Vocabulary V MX In vocabulary library V M The weight of (a);
S A (V MX ) -by vocabulary V MX The sub word bank V M The number of incidents caused by the vocabulary in (a) is increased.
The following formulas (1-10) are weight calculation methods of any word in the sub-class word stock.
V in Zx -any vocabulary in the lexicon V;
W(V Zx ) Vocabulary V Zx In word stockWeights in V;
S A (V Zx ) -by vocabulary V Zx Number of incidents caused by vocabulary in the sub-word library
S A (V Zx ) -total number of times the vocabulary in each sub-class word library causes an accident to occur
The purpose of setting the sub word stock weights is to perfect the extraction work of the accident data text risk feature words, consider the comprehensive importance degree of the risk feature words in the text set and the accident professional word stock, and eliminate the result errors caused by only considering the weights of the text set. The text set is accident data text extracted from a rail traffic accident report which occurs in a history.
The step S20 specifically includes a processing procedure for extracting a risk concept word of the accident data text provided in the embodiment of the present invention is shown in fig. 2. According to the present situation of unstructured storage of accident data texts in rail transit operation, firstly, word segmentation processing of the accident data texts is completed by utilizing a Chinese text word segmentation technology, and word segmentation effects of the accident data texts are optimized in a mode of combining hidden Markov and Viterbi algorithms, so that an accident word segmentation set is obtained. And obtaining the weight of each accident word according to the calculation method of the weight of any word in the sub-class word stock in the word stock. Traversing the accident professional word stock for each accident word, judging whether each accident word belongs to the four sub word stocks, and if so, determining that the accident word is a risk concept word of the accident data text. And then, comprehensively considering the weight of each sub word stock in the accident professional word stock and the weight of the vocabulary in each sub word stock in the whole accident professional word stock, calculating the weight value of the risk concept word of the accident data text by using a TF-IDF algorithm, and sequencing the weight values of the risk concept words.
In order to further define the occurrence mechanism of the operation accident, on the basis of extracting the risk concept words of the accident data text, confidence and extraction of the hierarchical and non-hierarchical relationship between the risk concept words around the top-level risk concept words based on an improved K-means algorithm are applied to complete connection between related risk concepts with hierarchical relationship and non-hierarchical relationship with the risk ontology top-level risk concept.
(1) Extraction of risk concept words of accident data text
The characteristics that the urban rail operation accident data text contains time, place, special punctuation and other information irrelevant to the research are comprehensively considered, so that a hidden Markov model with the comprehensive advantages of high data processing speed, high efficiency and the like is selected to complete the word segmentation of the accident data text, and the Viterbi algorithm model is combined to optimize the word segmentation result to realize the efficient and accurate word segmentation of the accident data text.
Because four sub word libraries are divided in the operation accident professional word library according to different weight values, the comprehensive calculation of the weight values of the risk concept words is realized by utilizing a TF-IDF algorithm by combining the weight of each sub word library in the accident professional word library and the specific gravity of the vocabulary in each sub word library in the whole accident professional word library. This way it is possible to embody the global importance of the selected risk concept words, i.e. "number" is not equivalent to "quality". Therefore, in order to effectively improve the extraction precision of the TF-IDF algorithm, the improvement work of the TF-IDF algorithm is completed by combining the constructed accident professional word stock library representing the accident characteristics. The calculation modes of the weight values of the risk concept words are formulas (1-11) and (1-12).
And, in addition, the method comprises the steps of,
δ+γ＝1 (1-12)
delta-coefficient factor for measuring comprehensive weight of risk feature words, wherein delta is more than 0 and less than 1;
gamma, a coefficient factor for measuring comprehensive weight of risk feature words, wherein gamma is more than 0 and less than 1.
Wherein: w (C) x ) Risk ontology concept word C x Weights of (2);
-some risk ontology concept word C x In the accident text d W The number of occurrences of (a);
accident text d W The total number of all the segmented words in the database;
i D i-total number of accident texts;
-some risk ontology concept word C x In accident text D W Is the number of occurrences.
(2) Extracting risk ontology risk concept relationship
The relationship among the risk concepts in the risk ontology model is divided into a longitudinal hierarchical relationship reflecting the risk concept hierarchy and a transverse non-hierarchical relationship reflecting the risk concept association. The longitudinal hierarchical relationship describes the upper and lower relationship and father-son relationship among the risk concepts, and indicates the subordinate relationship among the risks. For example: the shaft couplings, the axles, the wheels, the axle boxes and the like in the urban rail transit vehicle system are components of the bogie, and the shaft couplings, the axles, the wheels, the axle boxes and the bogie are described as having subordinate relations; in the risk ontology model, relationships between such risk concept words may be described in terms of hierarchical relationships. Non-hierarchical relationships refer to other connection relationships between risk concept words in addition to hierarchical relationships; in the construction of the risk ontology, non-hierarchical relations among risk concept words are utilized to determine sub-class risk concepts of the parent class risk concepts, namely, a preliminary framework of a risk ontology model is determined.
In the process of extracting concept words of the risk ontology of an operation accident, some words often appear simultaneously with other words, namely, the words and other words have tight connection relations, and the words are called as concept words of the top level of the risk ontology. Such top-level concept words have an important role in risk ontology research and therefore, it is necessary to determine such words.
In the selection process of the top concept words, the description follows a top-down modeling principle, and comprehensively considers the relation between the accident text set and the risk ontology concept words. The following formula is used for selecting the basis for the top concept words of the risk ontology:
k is the total amount of accident text;
m-total number of risk ontology concept words;
C x ,C y -representing arbitrary conceptual words;
f(C xi ,C yi ) Concept word C x ，C y The frequency of occurrence is the same as the ith accident text;
F(C x ,C y ) Concept word C x ，C y Co-occurrence frequency in accident text set.
After extracting the risk ontology concept words, further researching the interlayer relation among the risk ontology concept words is needed, and the calculation formula is as follows
In the formula, con (C) x ,C y ) Concept word C x ，C y Confidence between concepts used for representing interlayer relations between concepts;
F(C x ) -express concept word C x Co-occurrence frequency in accident text set
According to the improved K-means algorithm, a preliminary framework of a risk body of the rail transit operation accident is obtained, and the specific processing process is as follows:
step1: determining top-level risk concept words in a risk concept word set according to formulas (1-14), and fig. 3 is a schematic diagram of extracting top-level risk concept words and inter-word clustering provided by an embodiment of the invention.
Step2: establishing a confidence relation between the top-level risk concept words and other risk concept words according to the formula (1-15), and judging the number of sub-risk concept words of the top-level risk concept words;
step3: if the confidence level is greater than the threshold, i.e. Con (C t ,C x ) If the sub risk concept word is more than T, a connection relation exists between the sub risk concept word and the top risk concept word, and a space vector is constructed between the sub risk concept word and the top risk concept word as follows;
V(C i ,C x )＝[Con(C i ,C x ),W(C x )] (1-16)
wherein: v (C) i ,C x ) -top-level concept C i With its child node C x Is a spatial representation vector of (c). Step4: if the confidence level is less than the threshold, i.e. Con (C t ,C x ) No connection relation exists between the sub risk concept words and the top risk concept words if T is not more than or equal to T;
step5: fig. 4 is a schematic diagram of intra-cluster risk concept word screening, connection between each cluster and a top-level risk concept, and a risk concept word relationship structure provided in an embodiment of the present invention. Clustering by using an improved K-means algorithm to obtain a cluster with Z= { Z mi I=1, 2,3,..k } is a risk concept cluster of cluster centers, namely:
Cluster＝{Z 1 {C 1 ,C 2 ,...,C x },...,Z i {C 1 ,C 2 ,...,C y },...,Z k {C 1 ,C 2 ,...,C z }}；
step6: repeating the steps, and clustering each cluster until the cluster center is not changed any more;
step7: judging top-level risk concept words in each clustered risk concept cluster, and taking the top-level risk concept words as child node risk concept words of the top-level risk concept words of the upper layer;
step8: steps 3 through 7 are repeated until the determination of the relationships between all risk concept words is completed.
The step S30 specifically includes the steps of completing the construction of the risk ontology of the rail transit operation after the hierarchical and non-hierarchical relationship among the risk concept words of the rail transit operation and the related attributes such as the weight value of the risk concept words are defined based on the risk concept of the risk ontology of the rail transit operation.
(1) Risk ontology risk concept class and hierarchy structure for defining rail transit operation
The risk ontology structure can clearly present the hierarchical relationship and the non-hierarchical relationship among the risk concept words extracted based on the accident data text. On the basis that the research has determined the top-level risk concept of the risk ontology, the connection between related risk concepts with hierarchical relationship and non-hierarchical relationship with the top-level risk concept of the risk ontology is completed, and the preliminary construction of the risk ontology structure is realized; fig. 5 is a block diagram of a top-level risk concept class according to an embodiment of the present invention. A partial ontology description language for constructing the top-level risk concept class is shown in fig. 6 below.
(2) Risk ontology risk concept relationship
The construction of the risk ontology model is to essentially express the logical relations among the risk concept words representing the accident, wherein the logical relations among the risk concept words mainly represent the hierarchical relations, the parallel relations and the correlation relations representing the non-hierarchical relations; specific functional and ontology language descriptions are shown in the following table.
(3) Risk ontology risk concept attributes
And (3) completing structural description of the hierarchical and non-hierarchical relationship among risk concept words, namely completing construction of a hierarchical framework of the urban rail transit operation risk accident body. In order to further perfect the risk ontology model, the attribute of the risk concept word in the risk ontology needs to be represented, and the risk concept attribute is divided into an object attribute and a data attribute. The object attribute expresses the relation existing among the risk concept words, and the part is already elaborated in the risk concept class and the hierarchical relation links; in addition, the object attribute of each risk concept word in the risk ontology has its own definition domain and value domain, the definition domain and the value domain are used for performing formal constraint and specification on each risk concept individual, the description language of the risk ontology is respectively represented by Domains and Ranges, and meanwhile, the constraint characteristic still has its own reversibility and transitivity between levels. The object attribute represents a data pattern relationship between risk concept words, and generally includes a state value of the risk concept word, a conditional probability between risk concept words without hierarchy, a priori probability, and the like.
And according to the analysis, the construction of the urban rail transit operation accident risk ontology model is completed. Storage of risk knowledge based on accident data text extraction in a structured, formal ontology model schema can be achieved using OWL (Ontology Wed Language, ontology description language).
In summary, the embodiment of the invention discloses a method for constructing a rail transit system operation risk ontology, which mainly solves the following problems: in the current risk research process, the main data source is an accident report in a text form, so how to extract key accident information from the text report is of great importance. Therefore, from four aspects of man-machine loop, the invention establishes a professional word stock for representing the characteristics of system operation accidents and performs weight setting; an operation risk knowledge extraction method utilizing a hidden Markov and Viterbi algorithm and an improved K-means algorithm is provided; and constructing the rail transit system operation risk ontology according to the concepts of the risk ontology, the concept relationships and the attributes among the concepts.
The method can effectively extract the text risk of the urban rail transit operation accident, and meanwhile, in order to further utilize the risk related information in a limited way, the specification provides a risk ontology model capable of representing the operation accident, and the effective storage of risk knowledge is completed in a structured, formalized and standardized mode. The method provides important data support for the rail transit system operation risk research, and lays a solid foundation for the risk prevention and control research.
Those of ordinary skill in the art will appreciate that: the drawing is a schematic diagram of one embodiment and the modules or flows in the drawing are not necessarily required to practice the invention.
From the above description of embodiments, it will be apparent to those skilled in the art that the present invention may be implemented in software plus a necessary general hardware platform. Based on such understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a storage medium, such as a ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method described in the embodiments or some parts of the embodiments of the present invention.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments. In particular, for apparatus or system embodiments, since they are substantially similar to method embodiments, the description is relatively simple, with reference to the description of method embodiments in part. The apparatus and system embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.
The present invention is not limited to the above-mentioned embodiments, and any changes or substitutions that can be easily understood by those skilled in the art within the technical scope of the present invention are intended to be included in the scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.
Claims (4)
1. The construction method of the rail transit system operation risk ontology is characterized by comprising the following steps:
according to the operation accident characteristics of the rail transit system, starting from four aspects of a man-machine loop, constructing a five-class sub word library representing the accident characteristics, and setting weights of the five-class sub word library;
calculating risk concept words of the accident data text by utilizing improved TF-IDF based on the accident data text of the rail transit operation and the five-class sub word library, calculating weight values of the risk concept words according to weights corresponding to the five-class sub word library, and extracting hierarchical and non-hierarchical relations among the risk concept words by utilizing an improved K-means algorithm;
constructing a rail transit system operation risk ontology according to the risk concepts of the rail transit operation risk ontology, the hierarchical and non-hierarchical relations among the risk concept words and the weight values of the risk concept words;
the construction of the five-class sub word library for representing accident characteristics starts from four aspects of man-machine loop according to the operation accident characteristics of the rail transit system, and comprises the following steps:
from the accident feature of the rail transit system operation, dividing the accident professional word stock of the urban rail transit operation into four sub word stocks of personnel, physical component class, environment class and safety management class;
respectively using formulas (1-1), (1-2), (1-3), (1-4) and (1-5) to represent vocabulary sets in the accident professional word stock, the personnel sub word stock, the physical group classification sub word stock, the environment sub word stock and the safety management sub word stock;
V＝{V P , V S , V E , V M } (1-1)
V P ＝{V P1 , V P2 , V P3 , ..., V Pn } (1-2)
V S ＝{V S1 , V S2 , V S3 , ..., V Sm } (1-3)
V E ＝{V E1 , V E2 , V E3 , ..., V Ei } (1-4)
V M ＝{V M1 , V M2 , V M3 , ..., V Mj } (1-5)
v in P -a collection of staff-like vocabulary;
V S -a collection of physical group classification words;
V E -an environment-like vocabulary set;
V M -managing a collection of class vocabularies;
n-total number of personnel vocabulary;
m-total number of physical group classification vocabulary;
i-the total number of environmental vocabulary;
j-managing the total number of the class vocabularies;
the weight setting of the five-class sub word library comprises the following steps:
calculating the proportion of the vocabulary in each sub word stock in the whole accident professional word stock by using formulas (1-6), (1-7), (1-8) and (1-9) respectively;
assume that
Same reason
Same reason
Same reason
W(V PX ) Vocabulary V PX In vocabulary library V P The weight of (a);
S A (V PX ) -by vocabulary V PX The sub word bank V P The number of occurrences of the accident caused by the vocabulary in (a);
W(V SX ) Vocabulary V SX In vocabulary library V S The weight of (a);
S A (V SX ) -by vocabulary V SX The sub word bank V S The number of occurrences of the accident caused by the vocabulary in (a);
W(V EX ) Vocabulary V EX In vocabulary library V E The weight of (a);
S A (V EX ) -by vocabulary V EX The sub word bank V E The number of occurrences of the accident caused by the vocabulary in (a);
W(V MX ) Vocabulary V MX In vocabulary library V M The weight of (a);
S A (V MX ) -by vocabulary V MX The sub word bank V M The number of occurrences of the accident caused by the vocabulary in (a);
calculating the weight of any word in the sub-class word stock by using the formula (1-10);
and Z { P, S, E, M }, respectively
(1-10) formula V Zx -any vocabulary in the lexicon V;
W(V Zx v) -vocabulary V Zx Weights in word stock V;
S A (V Zx ) -by vocabulary V Zx Number of incidents caused by vocabulary in the sub-word library
S A (V) -the total number of incidents caused by the vocabulary in each sub-class word stock;
extracting hierarchical and non-hierarchical relations between risk concept words by using an improved K-means algorithm, wherein the method comprises the following steps:
dividing the relationship among the risk concept words into a longitudinal hierarchical relationship reflecting the risk concept hierarchy and a transverse non-hierarchical relationship reflecting the risk concept association in the risk ontology model, wherein the longitudinal hierarchical relationship describes the upper and lower relationship and father-son relationship among the risk concepts and indicates the subordinate relationship among the risks, and the non-hierarchical relationship refers to other connection relationships among the risk concept words except the hierarchical relationship;
the top-down modeling principle is followed, the top-level concept words are selected by comprehensively considering the relation between the accident text set and the risk ontology concept words, and the following formula is used for selecting the basis for the risk ontology top-level concept words:
k is the total amount of accident text;
m-total number of risk ontology concept words;
C x ,C y -representing arbitrary conceptual words;
f(C xi ,C yi ) Concept word C x ，C y The frequency of occurrence is the same as the ith accident text;
F(C x ,C y ) Concept word C x ，C y Co-occurrence frequency in accident text set;
the calculation formula of the interlayer relation among the risk ontology concept words is as follows:
in the formula, con (C) x ,C y ) Concept word C x ，C y Confidence between concepts used for representing interlayer relations between concepts;
F(C x ) -express concept word C x Co-occurrence frequency in accident text set;
the processing procedure for extracting the hierarchical and non-hierarchical relations between the risk concept words by using the improved K-means algorithm comprises the following steps:
step1: determining top-level risk concept words in the risk concept word set according to formulas (1-14);
step2: establishing a confidence relation between the top-level risk concept words and other risk concept words according to the formula (1-15), and judging the number of sub-risk concept words of the top-level risk concept words;
step3: if the confidence level is greater than the threshold, i.e. Con (C t ,G x ) Judging that a connection relationship exists between the sub-risk concept word and the top-level risk concept word, and constructing a space vector between the sub-risk concept word and the top-level risk concept word;
step4: if the confidence level is less than the threshold, i.e. Con (C t ,G x ) Judging that no connection relation exists between the sub risk concept words and the top risk concept words if T is less than or equal to T;
step5: clustering by using an improved K-means algorithm to obtain a cluster with Z= { Z mi I=1, 2,3,..k } is a risk concept cluster of cluster centers, namely:
Cluster＝{Z 1 {C 1 ,C 2 ,...,C x },...,Z i {C 1 ,C 2 ,...,C y },...,Z k {C 1 ,C 2 ,...,C z }}；
step6: repeating the steps, and clustering each cluster until the cluster center is not changed any more;
step7: judging top-level risk concept words in each clustered risk concept cluster, and taking the top-level risk concept words as child node risk concept words of the top-level risk concept words of the upper layer;
step8: steps 3 through 7 are repeated until the determination of the relationships between all risk concept words is completed.
2. The method of claim 1, wherein the accident data text based on the rail transit operation and the five-class sub word library calculate risk concept words of the accident data text by using improved TF-IDF, and calculate weight values of the risk concept words according to weights corresponding to the five-class sub word library, comprising:
performing word segmentation processing on the accident data text by using a Chinese text word segmentation technology, and optimizing word segmentation effect of the accident data text in a mode of combining hidden Markov and Viterbi algorithm to obtain an accident word segmentation set;
and respectively traversing the accident professional word stock for each accident word, judging whether each accident word belongs to the four sub word stocks, if so, determining that each accident word is a risk concept word of the accident data text, comprehensively considering the weight of each sub word stock in the accident professional word stock and the weight of the vocabulary in each sub word stock in the whole accident professional word stock, calculating the weight value of the risk concept word of the accident data text by utilizing a TF-IDF algorithm, and sequencing the weight values of the risk concept words.
3. The method of claim 2, wherein said comprehensively considering the weight of each sub word stock in the accident professional word stock and the weight of the vocabulary in each sub word stock in the whole accident professional word stock, calculating the weight value of the risk concept word of the accident data text by using TF-IDF algorithm, comprises:
the calculation modes of the weight values of the risk concept words are formulas (1-11) and (1-12);
and, in addition, the method comprises the steps of,
δ+γ＝1 (1-12)
delta-coefficient factor for measuring comprehensive weight of risk feature words, wherein delta is more than 0 and less than 1;
gamma, a coefficient factor for measuring comprehensive weight of risk feature words, wherein gamma is more than 0 and less than 1.
4. The method of claim 3, wherein the constructing the rail transit system operation risk ontology according to the risk concepts of the rail transit operation risk ontology, the hierarchical and non-hierarchical relationships between the risk concept words, and the weight values of the risk concept words comprises:
according to the hierarchical and non-hierarchical relationship between risk concept words, the connection between related risk concepts with hierarchical relationship and non-hierarchical relationship with the risk concept words on the top layer of the risk ontology is completed, and the preliminary construction of the operation risk ontology structure of the rail transit system is realized;
the method comprises the steps of completing structural description of hierarchical and non-hierarchical relations among risk concept words by representing attributes of the risk concept words in a risk ontology, and constructing a hierarchical framework of an operation risk ontology of a rail transit system, wherein the attributes of the risk concept words comprise object attributes and data attributes;
and according to the initially constructed rail transit system operation risk body structure and the hierarchical framework of the rail transit system operation risk body, the construction of the rail transit system operation risk body is completed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110791328.XA CN113535978B (en) | 2021-07-13 | 2021-07-13 | Construction method of rail transit system operation risk ontology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110791328.XA CN113535978B (en) | 2021-07-13 | 2021-07-13 | Construction method of rail transit system operation risk ontology |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113535978A CN113535978A (en) | 2021-10-22 |
| CN113535978B true CN113535978B (en) | 2024-03-19 |
Family
ID=78098916
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110791328.XA Active CN113535978B (en) | 2021-07-13 | 2021-07-13 | Construction method of rail transit system operation risk ontology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113535978B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115081918B (en) * | 2022-07-07 | 2023-05-23 | 北京交通大学 | Rail transit risk point prediction method and system based on data driving |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2006285825A (en) * | 2005-04-04 | 2006-10-19 | Hitachi Ltd | Support system for quantifying risk and support method for quantifying risk |
| CN104123609A (en) * | 2014-07-05 | 2014-10-29 | 华中科技大学 | Metro construction risk knowledge construction method based on noumenon |
| CN108492054A (en) * | 2018-04-11 | 2018-09-04 | 北京交通大学 | Urban track traffic risk chain group construction method based on multitiered network |
| CN112580998A (en) * | 2020-12-24 | 2021-03-30 | 北京交通大学 | Control method for rail transit operation risk |
-
2021
- 2021-07-13 CN CN202110791328.XA patent/CN113535978B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2006285825A (en) * | 2005-04-04 | 2006-10-19 | Hitachi Ltd | Support system for quantifying risk and support method for quantifying risk |
| CN104123609A (en) * | 2014-07-05 | 2014-10-29 | 华中科技大学 | Metro construction risk knowledge construction method based on noumenon |
| CN108492054A (en) * | 2018-04-11 | 2018-09-04 | 北京交通大学 | Urban track traffic risk chain group construction method based on multitiered network |
| CN112580998A (en) * | 2020-12-24 | 2021-03-30 | 北京交通大学 | Control method for rail transit operation risk |
Non-Patent Citations (1)
| Title |
|---|
| 王艳辉 ; 晋君 ; 李曼.《城轨路网下车站重要度的评价方法及应用》.《解放军理工大学学报》.2013,第1-7页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113535978A (en) | 2021-10-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113392986B (en) | Highway bridge information extraction method based on big data and management maintenance system | |
| Wang et al. | A hybrid document feature extraction method using latent Dirichlet allocation and word2vec | |
| CN113312411B (en) | Equipment fault cause diagnosis method based on knowledge graph and rule constraint | |
| Le et al. | Text classification: Naïve bayes classifier with sentiment Lexicon | |
| CN104346379A (en) | Method for identifying data elements on basis of logic and statistic technologies | |
| Dong et al. | Identifying incident causal factors to improve aviation transportation safety: Proposing a deep learning approach | |
| CN113535978B (en) | Construction method of rail transit system operation risk ontology | |
| CN110472225A (en) | The railway accident analysis of causes method of word-based extension LDA | |
| Wang et al. | Traffic information mining from social media based on the MC-LSTM-Conv model | |
| Wenhui et al. | Bayesian network-based knowledge graph inference for highway transportation safety risks | |
| CN115114455A (en) | Ontology-based multi-granularity urban rainstorm waterlogging knowledge map construction method | |
| CN114818719A (en) | Community topic classification method based on composite network and graph attention machine mechanism | |
| CN114547300A (en) | Relationship classification method combining remote supervision and supervised | |
| Zhang et al. | Similarity judgment of civil aviation regulations based on Doc2Vec deep learning algorithm | |
| CN112732787A (en) | Equipment portrait and personalized operation and maintenance service method for important parts of motor train unit | |
| CN113435190B (en) | Chapter relation extraction method integrating multilevel information extraction and noise reduction | |
| Jiang et al. | Towards end-to-end learning for efficient dialogue agent by modeling looking-ahead ability | |
| Yang et al. | Defect knowledge graph construction and application in multi-cloud IoT | |
| Wu et al. | An automatic text generation algorithm of technical disclosure for catenary construction based on knowledge element model | |
| Liu | Usability of Computer-Aided Translation Software Based on Deep Learning Algorithms | |
| Noursalehi et al. | Machine-learning-augmented analysis of textual data: application in transit disruption management | |
| Wang et al. | SMAAMA: A named entity alignment method based on Siamese network character feature and multi-attribute importance feature for Chinese civil aviation | |
| Wang et al. | Social networks and railway passenger capacity: An empirical study based on text mining and deep learning | |
| Hong et al. | Traffic flow prediction using spatiotemporal analysis and encoder-decoder network | |
| Wang et al. | Customized product configuration rule intelligent extraction and dynamic updating method based on the least recently used dynamic decision tree |
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 |