CN113592040A - Method and device for classifying dangerous chemical accidents - Google Patents
Method and device for classifying dangerous chemical accidents Download PDFInfo
- Publication number
- CN113592040A CN113592040A CN202111132131.1A CN202111132131A CN113592040A CN 113592040 A CN113592040 A CN 113592040A CN 202111132131 A CN202111132131 A CN 202111132131A CN 113592040 A CN113592040 A CN 113592040A
- Authority
- CN
- China
- Prior art keywords
- attribute
- accident
- attributes
- determining
- information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 239000000126 substance Substances 0.000 title claims abstract description 97
- 238000000034 method Methods 0.000 title claims abstract description 45
- 238000012549 training Methods 0.000 claims abstract description 97
- 238000013145 classification model Methods 0.000 claims abstract description 60
- 239000000383 hazardous chemical Substances 0.000 claims abstract description 46
- 238000012360 testing method Methods 0.000 claims description 32
- 238000013528 artificial neural network Methods 0.000 claims description 16
- 238000013139 quantization Methods 0.000 claims description 15
- 238000004590 computer program Methods 0.000 claims description 13
- 229910052739 hydrogen Inorganic materials 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 2
- 238000004458 analytical method Methods 0.000 abstract description 2
- 230000000875 corresponding effect Effects 0.000 description 57
- 230000008569 process Effects 0.000 description 7
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 description 6
- 238000012545 processing Methods 0.000 description 6
- 230000006870 function Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000004880 explosion Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 239000003345 natural gas Substances 0.000 description 3
- 238000003062 neural network model Methods 0.000 description 3
- 230000009467 reduction Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- 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
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Tourism & Hospitality (AREA)
- Educational Administration (AREA)
- Development Economics (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a method and a device for classifying dangerous chemical accidents, belonging to the field of dangerous chemical accident analysis, wherein the method comprises the following steps: determining a plurality of basic attributes related to the hazardous chemical accident; determining at least two associated attributes for influencing the classification result of the dangerous chemical accident from the plurality of basic attributes; training by utilizing at least two training sample pairs to obtain an accident classification model according to at least two correlation attributes; each training sample pair comprises corresponding sample dangerous chemical substance accident information and accident types; acquiring accident information of dangerous chemicals to be classified; extracting at least two correlation parameters from the information of the dangerous chemical accident to be classified according to at least two correlation attributes; the at least two associated parameters are in one-to-one correspondence with the at least two associated attributes; and obtaining the target accident category of the dangerous chemical substance accident to be classified by using the accident classification model and at least two associated parameters. According to the scheme, the reliability of the dangerous chemical accident classification result can be improved.
Description
Technical Field
The invention relates to the technical field of dangerous chemical accident analysis, in particular to a method and a device for classifying dangerous chemical accidents.
Background
Because dangerous chemicals have high potential threat, once relevant chemical accidents happen, huge casualties and property loss can be caused. Therefore, the accident reason classification result obtained by collecting and statistically analyzing the dangerous chemical accident information has important significance for researching accident rules, reducing the number of accidents, improving emergency handling capacity when the accidents happen and reducing loss caused by the accidents.
At present, classification of dangerous chemical substance accidents is realized by adopting a classifier, but when the classifier is used for analyzing and processing dangerous chemical substance accident information, the reliability of a classification result is poor.
Disclosure of Invention
The embodiment of the invention provides a method and a device for classifying dangerous chemical accidents.
In a first aspect, an embodiment of the present invention provides a method for classifying a hazardous chemical substance accident, including:
determining a plurality of basic attributes related to the hazardous chemical accident;
determining at least two associated attributes for influencing the classification result of the dangerous chemical accident from the plurality of basic attributes;
training by utilizing at least two training sample pairs to obtain an accident classification model according to the at least two correlation attributes; each training sample pair comprises corresponding sample dangerous chemical substance accident information and accident types;
acquiring accident information of dangerous chemicals to be classified;
extracting at least two correlation parameters from the dangerous chemical accident information to be classified according to the at least two correlation attributes; the at least two correlation parameters are in one-to-one correspondence with the at least two correlation attributes;
and obtaining the target accident category of the dangerous chemical substance accident to be classified by using the accident classification model and the at least two correlation parameters.
Preferably, the determining at least two associated attributes for influencing the classification result of the hazardous chemical accident from the plurality of basic attributes comprises:
for each of at least two training sample pairs, performing: extracting corresponding sample information parameters from the sample accident information in the training sample pair according to each basic attribute in the basic attributes to obtain a plurality of sample information parameters; the plurality of sample information parameters correspond to the plurality of basic attributes one to one;
combining the plurality of basic attributes to obtain a plurality of attribute groups, wherein each attribute group comprises at least two basic attributes;
for each attribute group of the plurality of attribute groups, performing: using each of the at least two training sample pairs, taking at least two sample information parameters of the training sample pair corresponding to at least two basic attributes included in the attribute group as the input of a BP neural network, taking the accident type in the training sample pair as the output of the BP neural network, and training the BP neural network to obtain an initial classification model corresponding to the attribute group; testing the initial classification model corresponding to the attribute group by using at least two test samples to obtain a test success rate; each test sample pair comprises corresponding sample hazardous chemical substance accident information and accident types;
determining a target attribute group corresponding to the maximum test success rate in the test success rates respectively corresponding to the plurality of attribute groups;
and determining each basic attribute included in the target attribute group as at least two associated attributes for influencing the classification result of the dangerous chemical accident.
Preferably, the obtaining an accident classification model by training with at least two training sample pairs according to the at least two correlation attributes includes: and determining an initial classification model corresponding to the target attribute grouping as the accident classification model.
Preferably, the determining at least two associated attributes for influencing the classification result of the hazardous chemical accident from the plurality of basic attributes comprises:
constructing a decision table by using the at least two training sample pairs; the decision table comprises a condition attribute and a decision attribute; the condition attributes are a set corresponding to each basic attribute in the training sample pairs, and the decision attributes are accident types in the training sample pairs;
calculating a first conditional information entropy of the conditional attribute to the decision attribute based on the decision table;
calculating a core attribute of the condition attribute relative to the decision attribute based on the decision table; determining the attribute corresponding to the difference value of the condition attribute and the core attribute as an intermediate set;
calculating a second condition information entropy of the current core attribute to the decision attribute, and judging whether the first condition information entropy is equal to the second condition information entropy or not; if not, calculating the importance of each attribute in the current middle set; determining a target attribute from the current intermediate set according to the importance of each attribute in the current intermediate set, deleting the target attribute from the current intermediate set, merging the target attribute into the current core attribute, and returning to the step of calculating the second conditional information entropy of the current core attribute on the decision attribute by using the merged current core attribute; and if so, determining each attribute in the current core attributes as at least two associated attributes for influencing the dangerous chemical accident classification result.
Preferably, the calculating the importance of each attribute in the current intermediate set includes: the importance of each attribute in the current intermediate set is calculated using the following equation:
SGF(Bi，P，D)=H(D|P)-H(D|P∪{Bi})
the SGF (Bi, P, D) is used to represent the importance of the ith attribute Bi in the current intermediate set B, H (dp | P) is used to represent the second condition information entropy of the current core attribute P to the decision attribute D, and H (dp | P ueq { Bi }) is used to represent the third condition information entropy of the union of the current core attribute P and the attribute Bi to the decision attribute D.
Preferably, the determining the target attribute from the current intermediate set according to the importance of each attribute in the current intermediate set includes: determining whether the number of attributes corresponding to the maximum importance in the current intermediate set is 1; if the importance degree is 1, determining the attribute corresponding to the maximum importance degree as the target attribute; otherwise, selecting one attribute from the attributes corresponding to the maximum importance degree to determine the attribute as the target attribute.
Preferably, the obtaining of the target accident category of the dangerous chemical substance accident to be classified by using the accident classification model and the at least two associated parameters includes:
quantizing the at least two associated parameters according to a set quantization logic to obtain at least two quantized values;
inputting the at least two quantized values into the accident classification model, and determining the output of the accident classification model as the target accident category.
In a second aspect, an embodiment of the present invention further provides a device for classifying a hazardous chemical accident, including:
a basic attribute determining unit for determining a plurality of basic attributes related to the hazardous chemical accident;
the relevant attribute determining unit is used for determining at least two relevant attributes for influencing the dangerous chemical accident classification result from the plurality of basic attributes;
the model training unit is used for utilizing at least two training sample pairs to train to obtain an accident classification model according to the at least two correlation attributes; each training sample pair comprises corresponding sample dangerous chemical substance accident information and accident types;
the accident information acquisition unit is used for acquiring accident information of the hazardous chemical substances to be classified;
the correlation parameter extraction unit is used for extracting at least two correlation parameters from the dangerous chemical accident information to be classified according to the at least two correlation attributes; the at least two correlation parameters are in one-to-one correspondence with the at least two correlation attributes;
and the accident category classification unit is used for obtaining the target accident category of the dangerous chemical substance accident to be classified by using the accident classification model and the at least two correlation parameters.
In a third aspect, an embodiment of the present invention further provides a computing device, including a memory and a processor, where the memory stores a computer program, and when the processor executes the computer program, the method for classifying a hazardous chemical substance accident according to any embodiment of this specification is implemented.
In a fourth aspect, the present invention further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed in a computer, the computer program causes the computer to execute the method for classifying a hazardous chemical accident according to any embodiment of the present specification.
The embodiment of the invention provides a method and a device for classifying dangerous chemical accidents.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
Fig. 1 is a flowchart of a method for classifying a hazardous chemical accident according to an embodiment of the present invention;
fig. 2 is a flowchart of an association attribute determining method according to an embodiment of the present invention;
fig. 3 is a flowchart of another method for determining an association attribute according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a decision table according to an embodiment of the present invention;
FIG. 5 is a diagram of a hardware architecture of a computing device according to an embodiment of the present invention;
fig. 6 is a structural diagram of an apparatus for classifying hazardous chemical substance accidents according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer and more complete, the technical solutions in the embodiments of the present invention will be described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention, and based on the embodiments of the present invention, all other embodiments obtained by a person of ordinary skill in the art without creative efforts belong to the scope of the present invention.
As mentioned above, classification of dangerous chemical accidents is currently implemented by using a classifier, and specifically, when analyzing and processing dangerous chemical accident information by using the classifier, it is generally necessary to extract information of specific attributes from the dangerous chemical accident information according to experience and then analyze the extracted information of the specific attributes, but since the specific attributes are obtained by human experience, some of the specific attributes may not be associated with the types of dangerous chemical accidents too much, and when classifying the dangerous chemical accident information by using the specific attributes, the obtained classification result has poor reliability. Based on the method, the correlation attributes which can influence the classification result of the dangerous chemical accident can be found out from the specific attributes, and the attributes which are not correlated with the classification result of the dangerous chemical accident are removed, so that the information of the dangerous chemical accident is analyzed by using the correlation attributes, and the reliability of the classification result is improved.
Specific implementations of the above concepts are described below.
Referring to fig. 1, an embodiment of the present invention provides a method for classifying a hazardous chemical accident, including:
102, determining at least two associated attributes for influencing the dangerous chemical accident classification result from the plurality of basic attributes;
104, training by using at least two training sample pairs to obtain an accident classification model according to the at least two correlation attributes; each training sample pair comprises corresponding sample dangerous chemical substance accident information and accident types;
108, extracting at least two correlation parameters from the information of the dangerous chemical accident to be classified according to the at least two correlation attributes; the at least two correlation parameters are in one-to-one correspondence with the at least two correlation attributes;
and 110, obtaining a target accident category of the dangerous chemical substance accident to be classified by using the accident classification model and the at least two correlation parameters.
In the embodiment of the invention, as the basic attributes possibly have the attributes irrelevant to the dangerous chemical accident classification result, the irrelevant attributes are removed, the relevant attributes for influencing the dangerous chemical accident classification result are reserved, and the accident classification model is trained by utilizing the relevant attributes, so that the output result of the accident classification model is more accurate, and the reliability of the dangerous chemical accident classification result is improved.
The manner in which the various steps shown in fig. 1 are performed is described below.
First, with respect to step 100, a plurality of basic attributes associated with a hazardous chemical incident are determined.
In one embodiment of the present invention, the basic attributes associated with the hazardous chemical accident may be obtained from empirical values, for example, the determined basic attributes may include: personnel composition, enterprise type, product inventory, safety standard annual training times, production equipment annual overhaul times, whether pipelines are broken or not, whether open fire exists or not in case of accidents or whether explosion is accompanied or not and the like.
Then, at least two related attributes for influencing the classification result of the dangerous chemical accident are determined from the plurality of basic attributes in step 102.
Assuming that the number of the basic attributes determined in step 100 is 9, and X1 to X9 are used to represent each basic attribute, it is necessary to determine the associated attributes that can affect the classification result of the dangerous chemical accident from among the 9 basic attributes, and considering that the reliability of the classification result of single attribute information is poor when analyzing the type of the dangerous chemical accident, at least two aspects of attribute information are required.
In one embodiment of the present invention, this step 102 can be determined at least in two ways:
the first mode is determined by utilizing the accuracy of the neural network model.
And secondly, reducing the attributes by using the information entropy.
The above two modes will be described below.
Referring to fig. 2, the first step 102 may include:
In one embodiment of the invention, each training sample pair comprises corresponding sample hazardous chemical substance accident information and accident type. The sample hazardous chemical substance accident information can be any one of structured information, semi-structured information and unstructured information. The accident category is the accident classification finally determined by the sample dangerous chemical accident information, such as natural gas leakage, open fire explosion and the like.
In an embodiment of the present invention, when extracting corresponding sample information parameters from sample accident information, if the sample accident information is semi-structured information or unstructured information, the sample accident information may be subjected to word segmentation processing, then each word obtained after the word segmentation processing is subjected to semantic understanding by using a natural language processing technology, and then the sample information parameters are extracted.
Taking one training sample pair as an example, the sample accident information in the training sample pair is extracted, and the following sample information parameters are extracted respectively according to the basic attributes X1-X9: the total number of people of the school calendar of the X1-personnel composition major experts is 40%, the X2-enterprise type is a production enterprise, the X3-product type is natural gas, the X4-product inventory is 20 tons, the X5-safety standard annual training frequency is 4 times, the X6-production equipment annual inspection and repair frequency is 3 times, an X7-pipeline is broken, open fire exists when an X8-accident occurs, and X9-is accompanied by explosion.
Step 202, combining the plurality of basic attributes to obtain a plurality of attribute groups, wherein each attribute group comprises at least two basic attributes.
In one embodiment of the present invention, combining multiple basic attributes may result in the following number of attribute groupings:
wherein N is used for the number of attribute groups, N is used for characterizing the number of basic attributes, N is an integer not less than 2,
is the number of combinations. When N =9, the above calculation results in N =502 attribute groups.
Taking one attribute group as an example, the basic attributes included in the attribute group are assumed to be X1-X7. And regarding each training sample pair, taking the sample information parameters of the basic attributes X1-X7 corresponding to the training sample pair as the input for training the BP neural network, and taking the accident category of the training sample pair as the output for training the BP neural network.
In an embodiment of the present invention, in order to facilitate training of the BP neural network, before training, the sample information parameters may be quantized according to a set quantization logic to obtain a quantization value, and the training of the BP neural network may be performed using the quantization value. For example, for the basic attribute X1, if the total number of people in the school calendar is 25% or less, the quantization value is 1, if the total number of people in the school calendar is 25% -60%, the quantization value is 2, and if the total number of people in the school calendar is 60% or more, the quantization value is 3. The quantization logic of the sample information parameters corresponding to other basic attributes can also be performed in this manner. For example, when the accident type is natural gas leakage, the quantization value is 1, and when the accident type is open fire explosion, the quantization value is 2.
An attribute group can be trained to obtain an initial classification model, after the initial classification model is obtained, the initial classification model is tested by using at least two test sample pairs, wherein each test sample pair comprises corresponding sample dangerous chemical accident information and an accident category, and when the initial classification model is tested, sample information parameters corresponding to basic attributes need to be extracted according to the sample dangerous chemical accident information of the test sample, for example, the basic attributes included in the attribute group corresponding to the initial classification model are X1-X7, so when the sample information parameters are extracted according to the test sample, the extracted sample information parameters need to correspond to X1-X7. And the quantization logic is also the same as that of the sample information parameter extracted from the training sample. Therefore, the parameters related in the training process and the testing process of the initial classification model are obtained according to the same processing logic, and the accuracy of the training result and the testing result is improved.
And step 206, determining the target attribute group corresponding to the maximum test success rate in the test success rates respectively corresponding to the plurality of attribute groups.
And step 208, determining each basic attribute included in the target attribute group as at least two associated attributes for influencing the dangerous chemical accident classification result.
In an embodiment of the present invention, when each initial classification model obtained by training is tested, the higher the test success rate is, the greater the influence of the basic attributes included in the attribute group on the accident classification result is, so that each basic attribute included in the target attribute group corresponding to the maximum test success rate can be determined as the associated attribute. For example, the determined association attributes are X2, X3, X4, X7, X8, and X9.
In the first implementation manner, the accuracy of the neural network model is used to determine the correlation attribute.
Next, the second embodiment will be explained.
Referring to the second method, referring to fig. 3, the step 102 may include:
In this step, before the decision table is constructed, sample information parameters may be extracted for each training sample pair, and specifically, the step 200 may be utilized to extract the sample information parameters of each training sample pair.
The decision table may be a table in which samples and their respective basic attributes are combined, and may be represented by S = (U, a = C ═ D, V, F). Wherein, U is used for characterizing the domain of discourse; c is used for representing the condition attribute and is a set of corresponding basic attributes in the training sample pair; d is decision attribute, including accident category in corresponding training sample pair;
the attribute evaluation device is used for representing a value set of each basic attribute; f: UxA → V is a mapping function, representing pairs
The following definitions are derived for the decision table: define 1, for T = (U, a = C = (U, C =) D, V, F), or expressed as K = (U, R), R is a set of equivalence relations of a non-empty set U (discourse domain), U/R represents a set of equivalence classes derived from the relation R, [ x =, [ x ] x]RRepresenting the R equivalence class containing the element x ∈ U.
Define 2, T = (U, a = C utoxyd, V, F), if Q is a subset of a and is not empty, there is a samplexi，xjFor any attribute C in the set Q, for an object on the domain of interesttAll have F (x)i，Ct)=F(xj，Ct) Then call xi，xjIndistinguishable (equivalent) is denoted ind (q).
Defining 3, P and Q as sets of equivalence relations on domain of discourse U, where Q is a subset of P, and Q is independent, with IND (P) = IND (Q), called P absolute reduction, the set of necessary relations in P called core CPRE (P) of P, the core of an equivalence relation cluster consisting of the intersection of all its reductions.
Define 4, C and D as equivalence relation cluster on domain U, and U/IND (C) = { x }1，x2，…，xn}，U/IND(D)={y1，y2，…，ynThe information entropy, i.e. uncertainty, of C can be expressed as:
the conditional information entropy of C relative to D can be expressed as:
for example, a decision table as shown in fig. 4 can be constructed by using the training sample pairs. Where m is the total number of training sample pairs, Y1, Y2 are different accident categories, respectively, and when Y1=1 and Y2=0, it indicates the accident category Y1.
In one embodiment of the present invention, the conditional information entropy of the conditional attribute C to the decision attribute D can be calculated using the conditional information entropy formula defined in step 300 as 4
In one embodiment of the invention, the core attributes may be computed using the following steps:
s1, order
Then calculating the conditional information entropy of the conditional attribute C based on the decision attribute D
S2, selecting the maximum condition information entropy
And calculate
S3, if
And deleted in the set C
Otherwise, go to S2 to continue execution until the condition attribute C is empty.
And calculating to obtain a core attribute CS, and if P = CS, then the intermediate set B is B = C-CS.
Wherein the second condition information entropy is
In one embodiment of the present invention, in the step 306, when calculating the importance of each attribute in the current intermediate set, the importance of each attribute in the current intermediate set can be calculated by using the following formula:
SGF(Bi，P，D)=H(D|P)-H(D|P∪{Bi})
the SGF (Bi, P, D) is used to represent the importance of the ith attribute Bi in the current intermediate set B, H (dp | P) is used to represent the second condition information entropy of the current core attribute P to the decision attribute D, and H (dp | P ueq { Bi }) is used to represent the third condition information entropy of the union of the current core attribute P and the attribute Bi to the decision attribute D.
It should be noted that the importance of each attribute may be calculated by other methods, such as calculation using information entropy, besides the above formula.
In an embodiment of the present invention, when determining the target attribute from the current intermediate set according to the importance of each attribute in the current intermediate set in step 306, the method may include: determining whether the number of attributes corresponding to the maximum importance in the current intermediate set is 1; if the importance degree is 1, determining the attribute corresponding to the maximum importance degree as the target attribute; otherwise, selecting one attribute from the attributes corresponding to the maximum importance degree to determine the attribute as the target attribute. One attribute is selected from the attributes corresponding to the maximum importance and determined as the target attribute, and the attribute with the minimum combination number with the attribute value of P may be selected as the target attribute, or one attribute may be randomly selected as the target attribute.
If it is calculated that an attribute with an importance level of 0 exists in the current intermediate set B, it is necessary to directly perform a deletion operation on the attribute with the importance level of 0.
Assuming that P = { X2, X3, X4, X7, X8, X9} can be derived from the decision table, the associated attributes are X2, X3, X4, X7, X8, and X9.
In the second mode, the attributes with high importance are sequentially merged into the set P until all information in the domain of discourse U can be represented in the set P, the basic attributes contained in the set P have the largest influence on the classification result of the dangerous chemical accident, the method realizes feature dimension reduction, reduces the complexity among data, only retains the attributes influencing the classification result, is beneficial to accelerating the training speed of the accident classification model, improves the performance of the accident classification model, and can effectively prevent overfitting.
Next, in step 104, according to the at least two correlation attributes, an accident classification model is obtained by training with at least two training sample pairs; each training sample pair comprises corresponding sample hazardous chemical substance accident information and accident types.
In this step 104, if the first method is used to determine the associated attributes in step 102, in this step 104, the initial classification model corresponding to the target attribute group may be directly determined as the accident classification model, and retraining is not required, so that the acquisition speed of the accident classification model is increased.
In this step 104, if the second mode is used to determine the correlation attributes in step 102, in this step 104, it is necessary to extract sample information parameters according to each correlation attribute for each training sample pair, and train the BP neural network model by using the sample information parameters of the training sample pairs, where the training mode is the same as the training mode shown in fig. 2, and the training mode shown in fig. 2 may be referred to, which is not described in detail in this embodiment.
Finally, aiming at the step 106 of acquiring accident information of the dangerous chemical substances to be classified, and the step 108 of extracting at least two associated parameters from the accident information of the dangerous chemical substances to be classified according to the at least two associated attributes; and step 110, obtaining a target accident category of the dangerous chemical substance accident to be classified by utilizing the accident classification model and the at least two correlation parameters, and simultaneously explaining the at least two correlation parameters in one-to-one correspondence with the at least two correlation attributes.
In an embodiment of the present invention, if the information input by the accident classification model in the training process is a quantized value obtained from the sample information parameter according to a set quantization logic, in this step, the at least two associated parameters need to be quantized according to the same quantization logic to obtain at least two quantized values, and then the at least two quantized values are input into the accident classification model, so as to determine the output of the accident classification model as the target accident category.
The accident classification model is obtained by utilizing the associated attribute training, so that the classification accuracy of the accident classification model is high, the reliability of the accident category result obtained by utilizing the accident classification model is high, and the classification speed is high.
As shown in fig. 5 and 6, the embodiment of the invention provides a device for classifying dangerous chemical accidents. The device embodiments may be implemented by software, or by hardware, or by a combination of hardware and software. In terms of hardware, as shown in fig. 5, for a hardware architecture diagram of a computing device where an apparatus for classifying hazardous chemical substance accidents according to an embodiment of the present invention is located, in addition to the processor, the memory, the network interface, and the nonvolatile memory shown in fig. 5, the computing device where the apparatus is located in the embodiment may generally include other hardware, such as a forwarding chip responsible for processing a packet, and the like. Taking a software implementation as an example, as shown in fig. 6, as a logical apparatus, a CPU of a computing device in which the apparatus is located reads a corresponding computer program in a non-volatile memory into a memory to run. The device that the embodiment provided classifies to dangerous chemicals accident includes:
a basic attribute determining unit 601, configured to determine a plurality of basic attributes related to a hazardous chemical accident;
an associated attribute determining unit 602, configured to determine at least two associated attributes for influencing the hazardous chemical substance accident classification result from the plurality of basic attributes;
the model training unit 603 is configured to obtain an accident classification model by training with at least two training sample pairs according to the at least two correlation attributes; each training sample pair comprises corresponding sample dangerous chemical substance accident information and accident types;
an accident information obtaining unit 604, configured to obtain accident information of hazardous chemical substances to be classified;
the associated parameter extracting unit 605 is configured to extract at least two associated parameters from the hazardous chemical substance accident information to be classified according to the at least two associated attributes; the at least two correlation parameters are in one-to-one correspondence with the at least two correlation attributes;
and an accident category classification unit 606, configured to obtain a target accident category of the dangerous chemical substance accident to be classified by using the accident classification model and the at least two correlation parameters.
In an embodiment of the present invention, the association attribute determining unit 602 is specifically configured to:
for each of at least two training sample pairs, performing: extracting corresponding sample information parameters from the sample accident information in the training sample pair according to each basic attribute in the basic attributes to obtain a plurality of sample information parameters; the plurality of sample information parameters correspond to the plurality of basic attributes one to one;
combining the plurality of basic attributes to obtain a plurality of attribute groups, wherein each attribute group comprises at least two basic attributes;
for each attribute group of the plurality of attribute groups, performing: using each of the at least two training sample pairs, taking at least two sample information parameters of the training sample pair corresponding to at least two basic attributes included in the attribute group as the input of a BP neural network, taking the accident type in the training sample pair as the output of the BP neural network, and training the BP neural network to obtain an initial classification model corresponding to the attribute group; testing the initial classification model corresponding to the attribute group by using at least two test samples to obtain a test success rate; each test sample pair comprises corresponding sample hazardous chemical substance accident information and accident types;
determining a target attribute group corresponding to the maximum test success rate in the test success rates respectively corresponding to the plurality of attribute groups;
and determining each basic attribute included in the target attribute group as at least two associated attributes for influencing the classification result of the dangerous chemical accident.
In an embodiment of the present invention, the model training unit 603 is specifically configured to determine an initial classification model corresponding to the target attribute group as the accident classification model.
In an embodiment of the present invention, the association attribute determining unit 602 is specifically configured to:
constructing a decision table by using the at least two training sample pairs; the decision table comprises a condition attribute and a decision attribute; the condition attributes are a set corresponding to each basic attribute in the training sample pairs, and the decision attributes are accident types in the training sample pairs;
calculating a first conditional information entropy of the conditional attribute to the decision attribute based on the decision table;
calculating a core attribute of the condition attribute relative to the decision attribute based on the decision table; determining the attribute corresponding to the difference value of the condition attribute and the core attribute as an intermediate set;
calculating a second condition information entropy of the current core attribute to the decision attribute, and judging whether the first condition information entropy is equal to the second condition information entropy or not; if not, calculating the importance of each attribute in the current middle set; determining a target attribute from the current intermediate set according to the importance of each attribute in the current intermediate set, deleting the target attribute from the current intermediate set, merging the target attribute into the current core attribute, and returning to the step of calculating the second conditional information entropy of the current core attribute on the decision attribute by using the merged current core attribute; and if so, determining each attribute in the current core attributes as at least two associated attributes for influencing the dangerous chemical accident classification result.
In an embodiment of the present invention, when performing the calculating of the importance of each attribute in the current intermediate set, the associated attribute determining unit 602 is specifically configured to calculate the importance of each attribute in the current intermediate set by using the following formula:
SGF(Bi，P，D)=H(D|P)-H(D|P∪{Bi})
the SGF (Bi, P, D) is used to represent the importance of the ith attribute Bi in the current intermediate set B, H (dp | P) is used to represent the second condition information entropy of the current core attribute P to the decision attribute D, and H (dp | P ueq { Bi }) is used to represent the third condition information entropy of the union of the current core attribute P and the attribute Bi to the decision attribute D.
In an embodiment of the present invention, when the associated attribute determining unit 602 determines the target attribute from the current intermediate set according to the importance of each attribute in the current intermediate set, it is specifically configured to determine whether the number of attributes corresponding to the maximum importance in the current intermediate set is 1; if the importance degree is 1, determining the attribute corresponding to the maximum importance degree as the target attribute; otherwise, selecting one attribute from the attributes corresponding to the maximum importance degree to determine the attribute as the target attribute.
In an embodiment of the present invention, the accident category classification unit 606 is specifically configured to quantize the at least two associated parameters according to a set quantization logic to obtain at least two quantized values; inputting the at least two quantized values into the accident classification model, and determining the output of the accident classification model as the target accident category.
It is to be understood that the illustrated structure of the embodiments of the present invention does not constitute a specific limitation on an apparatus for classifying hazardous chemical accidents. In other embodiments of the invention, an apparatus for classifying hazardous chemical accidents may include more or fewer components than shown, or some components may be combined, some components may be separated, or a different arrangement of components. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.
Because the content of information interaction, execution process, and the like among the modules in the device is based on the same concept as the method embodiment of the present invention, specific content can be referred to the description in the method embodiment of the present invention, and is not described herein again.
The embodiment of the invention also provides a computing device, which comprises a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to realize the method for classifying the dangerous chemical substance accidents in any embodiment of the invention.
An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, causes the processor to execute a method for classifying a hazardous chemical accident in any embodiment of the present invention.
Specifically, a system or an apparatus equipped with a storage medium on which software program codes that realize the functions of any of the above-described embodiments are stored may be provided, and a computer (or a CPU or MPU) of the system or the apparatus is caused to read out and execute the program codes stored in the storage medium.
In this case, the program code itself read from the storage medium can realize the functions of any of the above-described embodiments, and thus the program code and the storage medium storing the program code constitute a part of the present invention.
Examples of the storage medium for supplying the program code include a floppy disk, a hard disk, a magneto-optical disk, an optical disk (e.g., CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD + RW), a magnetic tape, a nonvolatile memory card, and a ROM. Alternatively, the program code may be downloaded from a server computer via a communications network.
Further, it should be clear that the functions of any one of the above-described embodiments may be implemented not only by executing the program code read out by the computer, but also by causing an operating system or the like operating on the computer to perform a part or all of the actual operations based on instructions of the program code.
Further, it is to be understood that the program code read out from the storage medium is written to a memory provided in an expansion board inserted into the computer or to a memory provided in an expansion module connected to the computer, and then causes a CPU or the like mounted on the expansion board or the expansion module to perform part or all of the actual operations based on instructions of the program code, thereby realizing the functions of any of the above-described embodiments.
It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an …" does not exclude the presence of other similar elements in a process, method, article, or apparatus that comprises the element.
Those of ordinary skill in the art will understand that: all or part of the steps for realizing the method embodiments can be completed by hardware related to program instructions, the program can be stored in a computer readable storage medium, and the program executes the steps comprising the method embodiments when executed; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A method of classifying a hazardous chemical accident, comprising:
determining a plurality of basic attributes related to the hazardous chemical accident;
determining at least two associated attributes for influencing the classification result of the dangerous chemical accident from the plurality of basic attributes;
training by utilizing at least two training sample pairs to obtain an accident classification model according to the at least two correlation attributes; each training sample pair comprises corresponding sample dangerous chemical substance accident information and accident types;
acquiring accident information of dangerous chemicals to be classified;
extracting at least two correlation parameters from the dangerous chemical accident information to be classified according to the at least two correlation attributes; the at least two correlation parameters are in one-to-one correspondence with the at least two correlation attributes;
and obtaining the target accident category of the dangerous chemical substance accident to be classified by using the accident classification model and the at least two correlation parameters.
2. The method for classifying a hazardous chemical accident according to claim 1, wherein the determining at least two associated attributes from the plurality of basic attributes for influencing the classification result of the hazardous chemical accident comprises:
for each of at least two training sample pairs, performing: extracting corresponding sample information parameters from the sample accident information in the training sample pair according to each basic attribute in the basic attributes to obtain a plurality of sample information parameters; the plurality of sample information parameters correspond to the plurality of basic attributes one to one;
combining the plurality of basic attributes to obtain a plurality of attribute groups, wherein each attribute group comprises at least two basic attributes;
for each attribute group of the plurality of attribute groups, performing: using each of the at least two training sample pairs, taking at least two sample information parameters of the training sample pair corresponding to at least two basic attributes included in the attribute group as the input of a BP neural network, taking the accident type in the training sample pair as the output of the BP neural network, and training the BP neural network to obtain an initial classification model corresponding to the attribute group; testing the initial classification model corresponding to the attribute group by using at least two test samples to obtain a test success rate; each test sample pair comprises corresponding sample hazardous chemical substance accident information and accident types;
determining a target attribute group corresponding to the maximum test success rate in the test success rates respectively corresponding to the plurality of attribute groups;
and determining each basic attribute included in the target attribute group as at least two associated attributes for influencing the classification result of the dangerous chemical accident.
3. The method for classifying hazardous chemical substance accidents according to claim 2, wherein the training with at least two training sample pairs according to the at least two correlation attributes to obtain an accident classification model comprises: and determining an initial classification model corresponding to the target attribute grouping as the accident classification model.
4. The method for classifying a hazardous chemical accident according to claim 1, wherein the determining at least two associated attributes from the plurality of basic attributes for influencing the classification result of the hazardous chemical accident comprises:
constructing a decision table by using the at least two training sample pairs; the decision table comprises a condition attribute and a decision attribute; the condition attributes are a set corresponding to each basic attribute in the training sample pairs, and the decision attributes are accident types in the training sample pairs;
calculating a first conditional information entropy of the conditional attribute to the decision attribute based on the decision table;
calculating a core attribute of the condition attribute relative to the decision attribute based on the decision table; determining the attribute corresponding to the difference value of the condition attribute and the core attribute as an intermediate set;
calculating a second condition information entropy of the current core attribute to the decision attribute, and judging whether the first condition information entropy is equal to the second condition information entropy or not; if not, calculating the importance of each attribute in the current middle set; determining a target attribute from the current intermediate set according to the importance of each attribute in the current intermediate set, deleting the target attribute from the current intermediate set, merging the target attribute into the current core attribute, and returning to the step of calculating the second conditional information entropy of the current core attribute on the decision attribute by using the merged current core attribute; and if so, determining each attribute in the current core attributes as at least two associated attributes for influencing the dangerous chemical accident classification result.
5. The method for classifying a hazardous chemical accident according to claim 4, wherein the calculating the importance of each attribute in the current intermediate set comprises: the importance of each attribute in the current intermediate set is calculated using the following equation:
SGF(Bi，P，D)=H(D|P)-H(D|P∪{Bi})
the SGF (Bi, P, D) is used to represent the importance of the ith attribute Bi in the current intermediate set B, H (dp | P) is used to represent the second condition information entropy of the current core attribute P to the decision attribute D, and H (dp | P ueq { Bi }) is used to represent the third condition information entropy of the union of the current core attribute P and the attribute Bi to the decision attribute D.
6. The method for classifying a hazardous chemical accident according to claim 4, wherein the determining a target attribute from the current intermediate set according to the importance of each attribute in the current intermediate set comprises: determining whether the number of attributes corresponding to the maximum importance in the current intermediate set is 1; if the importance degree is 1, determining the attribute corresponding to the maximum importance degree as the target attribute; otherwise, selecting one attribute from the attributes corresponding to the maximum importance degree to determine the attribute as the target attribute.
7. The method for classifying hazardous chemical substance accidents according to any one of claims 1 to 6, wherein the obtaining of the target accident category of the hazardous chemical substance accident to be classified by using the accident classification model and the at least two correlation parameters comprises:
quantizing the at least two associated parameters according to a set quantization logic to obtain at least two quantized values;
inputting the at least two quantized values into the accident classification model, and determining the output of the accident classification model as the target accident category.
8. A device for classifying hazardous chemical accidents, comprising:
a basic attribute determining unit for determining a plurality of basic attributes related to the hazardous chemical accident;
the relevant attribute determining unit is used for determining at least two relevant attributes for influencing the dangerous chemical accident classification result from the plurality of basic attributes;
the model training unit is used for utilizing at least two training sample pairs to train to obtain an accident classification model according to the at least two correlation attributes; each training sample pair comprises corresponding sample dangerous chemical substance accident information and accident types;
the accident information acquisition unit is used for acquiring accident information of the hazardous chemical substances to be classified;
the correlation parameter extraction unit is used for extracting at least two correlation parameters from the dangerous chemical accident information to be classified according to the at least two correlation attributes; the at least two correlation parameters are in one-to-one correspondence with the at least two correlation attributes;
and the accident category classification unit is used for obtaining the target accident category of the dangerous chemical substance accident to be classified by using the accident classification model and the at least two correlation parameters.
9. A computing device comprising a memory having stored therein a computer program and a processor that, when executing the computer program, implements a method of classifying a hazardous chemical accident according to any one of claims 1-7.
10. A computer-readable storage medium, on which a computer program is stored which, when executed in a computer, causes the computer to carry out the method of classifying a hazardous chemical accident of any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111132131.1A CN113592040B (en) | 2021-09-27 | 2021-09-27 | Method and device for classifying dangerous chemical accidents |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111132131.1A CN113592040B (en) | 2021-09-27 | 2021-09-27 | Method and device for classifying dangerous chemical accidents |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113592040A true CN113592040A (en) | 2021-11-02 |
| CN113592040B CN113592040B (en) | 2021-12-03 |
Family
ID=78242323
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111132131.1A Active CN113592040B (en) | 2021-09-27 | 2021-09-27 | Method and device for classifying dangerous chemical accidents |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113592040B (en) |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105930347A (en) * | 2016-04-05 | 2016-09-07 | 浙江远传信息技术股份有限公司 | Text analysis based power outage cause recognition system |
| CN108830767A (en) * | 2018-05-21 | 2018-11-16 | 尹云镭 | Accident vehicle travel speed and accident origin cause of formation aided analysis method |
| CN109523786A (en) * | 2018-11-30 | 2019-03-26 | 哈尔滨工业大学 | A kind of traffic accidents genetic analysis method |
| KR20190075392A (en) * | 2017-12-21 | 2019-07-01 | 한림대학교 산학협력단 | Classification system for detecting components of harmful chemical materials for quick response system according to chemical accident |
| CN110033011A (en) * | 2018-12-14 | 2019-07-19 | 阿里巴巴集团控股有限公司 | Traffic accident Accident Handling Method and device, electronic equipment |
| CN110110085A (en) * | 2019-04-24 | 2019-08-09 | 中电海康集团有限公司 | Traffic accident file classification method and system based on character level neural network and SVM |
| CN110134963A (en) * | 2019-05-20 | 2019-08-16 | 中南大学 | A kind of text mining is applied to the method for road traffic accident data processing |
| CN111209472A (en) * | 2019-12-24 | 2020-05-29 | 中国铁道科学研究院集团有限公司电子计算技术研究所 | Railway accident fault association and accident fault reason analysis method and system |
| CN111259149A (en) * | 2020-01-19 | 2020-06-09 | 清华大学 | Chemical accident classification method and device, computer equipment and storage medium |
| CN111275329A (en) * | 2020-01-20 | 2020-06-12 | 交通运输部水运科学研究所 | Water transport engineering construction safety accident cause analysis method |
| CN112183979A (en) * | 2020-09-18 | 2021-01-05 | 浙江省安全生产科学研究院 | Hazardous chemical substance loading and unloading safety risk assessment method based on multi-source information fusion |
| KR20210103361A (en) * | 2020-02-13 | 2021-08-23 | 울산대학교 산학협력단 | Classification system and method for wastes generated after chemical accident treatment |
-
2021
- 2021-09-27 CN CN202111132131.1A patent/CN113592040B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105930347A (en) * | 2016-04-05 | 2016-09-07 | 浙江远传信息技术股份有限公司 | Text analysis based power outage cause recognition system |
| KR20190075392A (en) * | 2017-12-21 | 2019-07-01 | 한림대학교 산학협력단 | Classification system for detecting components of harmful chemical materials for quick response system according to chemical accident |
| CN108830767A (en) * | 2018-05-21 | 2018-11-16 | 尹云镭 | Accident vehicle travel speed and accident origin cause of formation aided analysis method |
| CN109523786A (en) * | 2018-11-30 | 2019-03-26 | 哈尔滨工业大学 | A kind of traffic accidents genetic analysis method |
| CN110033011A (en) * | 2018-12-14 | 2019-07-19 | 阿里巴巴集团控股有限公司 | Traffic accident Accident Handling Method and device, electronic equipment |
| CN110110085A (en) * | 2019-04-24 | 2019-08-09 | 中电海康集团有限公司 | Traffic accident file classification method and system based on character level neural network and SVM |
| CN110134963A (en) * | 2019-05-20 | 2019-08-16 | 中南大学 | A kind of text mining is applied to the method for road traffic accident data processing |
| CN111209472A (en) * | 2019-12-24 | 2020-05-29 | 中国铁道科学研究院集团有限公司电子计算技术研究所 | Railway accident fault association and accident fault reason analysis method and system |
| CN111259149A (en) * | 2020-01-19 | 2020-06-09 | 清华大学 | Chemical accident classification method and device, computer equipment and storage medium |
| CN111275329A (en) * | 2020-01-20 | 2020-06-12 | 交通运输部水运科学研究所 | Water transport engineering construction safety accident cause analysis method |
| KR20210103361A (en) * | 2020-02-13 | 2021-08-23 | 울산대학교 산학협력단 | Classification system and method for wastes generated after chemical accident treatment |
| CN112183979A (en) * | 2020-09-18 | 2021-01-05 | 浙江省安全生产科学研究院 | Hazardous chemical substance loading and unloading safety risk assessment method based on multi-source information fusion |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113592040B (en) | 2021-12-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108737406B (en) | Method and system for detecting abnormal flow data | |
| CN111881289B (en) | Training method of classification model, and detection method and device of data risk class | |
| CN112511546A (en) | Vulnerability scanning method, device, equipment and storage medium based on log analysis | |
| CN109918648B (en) | Rumor depth detection method based on dynamic sliding window feature score | |
| CN113283675B (en) | Index data analysis method, device, equipment and storage medium | |
| CN110619535B (en) | Data processing method and device | |
| CN111221960A (en) | Text detection method, similarity calculation method, model training method and device | |
| CN113177643A (en) | Automatic modeling system based on big data | |
| CN114116973A (en) | Multi-document text duplicate checking method, electronic equipment and storage medium | |
| CN105323763B (en) | A kind of recognition methods of junk short message and device | |
| CN115063035A (en) | Customer evaluation method, system, equipment and storage medium based on neural network | |
| CN113590764B (en) | Training sample construction method and device, electronic equipment and storage medium | |
| CN113132368B (en) | Chat data auditing method and device and computer equipment | |
| CN111680120B (en) | News category detection method and system | |
| CN111723182B (en) | Key information extraction method and device for vulnerability text | |
| CN113592040B (en) | Method and device for classifying dangerous chemical accidents | |
| CN112990443A (en) | Neural network evaluation method and device, electronic device, and storage medium | |
| CN112016317A (en) | Sensitive word recognition method and device based on artificial intelligence and computer equipment | |
| CN111639494A (en) | Case affair relation determining method and system | |
| Garcia de Alford et al. | Reducing age bias in machine learning: An algorithmic approach | |
| CN113761875B (en) | Event extraction method and device, electronic equipment and storage medium | |
| CN112632229A (en) | Text clustering method and device | |
| CN114610576A (en) | Log generation monitoring method and device | |
| CN113420127A (en) | Threat information processing method, device, computing equipment and storage medium | |
| CN111859896B (en) | Formula document detection method and device, computer readable medium and electronic equipment |
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 |