CN110232121B - Semantic network-based control instruction classification method - Google Patents
Semantic network-based control instruction classification method Download PDFInfo
- Publication number
- CN110232121B CN110232121B CN201910348919.2A CN201910348919A CN110232121B CN 110232121 B CN110232121 B CN 110232121B CN 201910348919 A CN201910348919 A CN 201910348919A CN 110232121 B CN110232121 B CN 110232121B
- Authority
- CN
- China
- Prior art keywords
- control
- neural network
- verb
- control instruction
- argument
- 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
- 238000000034 method Methods 0.000 title claims abstract description 26
- 238000013528 artificial neural network Methods 0.000 claims abstract description 41
- 230000033228 biological regulation Effects 0.000 claims abstract description 19
- 238000002372 labelling Methods 0.000 claims abstract description 13
- 238000012545 processing Methods 0.000 claims abstract description 13
- 239000013598 vector Substances 0.000 claims description 31
- 230000001105 regulatory effect Effects 0.000 claims description 25
- 238000012549 training Methods 0.000 claims description 17
- 239000011159 matrix material Substances 0.000 claims description 13
- 230000011218 segmentation Effects 0.000 claims description 10
- 230000004913 activation Effects 0.000 claims description 6
- 210000002569 neuron Anatomy 0.000 claims description 6
- 239000012634 fragment Substances 0.000 claims 4
- 238000003062 neural network model Methods 0.000 claims 1
- 238000007781 pre-processing Methods 0.000 claims 1
- 230000009471 action Effects 0.000 abstract description 13
- 230000006870 function Effects 0.000 abstract description 6
- 238000000605 extraction Methods 0.000 abstract description 5
- 230000008569 process Effects 0.000 description 5
- 230000008901 benefit Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 239000000284 extract Substances 0.000 description 3
- 238000003058 natural language processing Methods 0.000 description 3
- 238000010276 construction Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000033001 locomotion Effects 0.000 description 2
- 239000003550 marker Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000013145 classification model Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3343—Query execution using phonetics
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
- G06F40/211—Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/237—Lexical tools
- G06F40/242—Dictionaries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- 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)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- Business, Economics & Management (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Databases & Information Systems (AREA)
- Computing Systems (AREA)
- Acoustics & Sound (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Machine Translation (AREA)
Abstract
The invention discloses a control instruction classification method based on a semantic network, aiming at: a computer-readable structured control instruction is constructed, and a basis is provided for the automatic processing of the control instruction; and forming a predicate logical structure to provide a basis for realizing a knowledge reasoning system. By processing the unstructured regulation instruction, the method can realize the following auxiliary functions: extracting information carried by basic control terms appearing in the control instruction for extraction; extracting and simulating the information of the aircraft such as the action and the state; computer-readable structured information is formed to provide data for knowledge reasoning. Aiming at the condition that a complex control instruction comprises a plurality of control intentions, the method judges whether different arguments in the control instruction are associated with different control intention verbs through a deep neural network, and brings the associated verbs and argument words into a semantic network for semantic role labeling.
Description
Technical Field
The invention belongs to the technical field of air traffic control automation systems, and particularly relates to a control instruction classification method based on a semantic network.
Background
With the vigorous development of the civil aviation industry in the last 30 years, the requirements of air traffic management are continuously expanded, so that the potential safety hazard problem is increasingly prominent. Statistics show that human factors account for over 75% of the past flight safety incidents, and among them, incidents due to controller error account for 25%. The mainstream method for solving the conflict caused by the error of the controller at present is to strengthen the monitoring equipment of the scene, and prevent the conflict from occurring by monitoring equipment such as a radar, a multipoint positioning system sensor and the like by the scene. Meanwhile, more advanced solutions based on artificial intelligence are also proposed, such as recognizing the control voice by using a voice recognition technology and converting the control voice into a text format, and extracting key information of the control instruction by using a natural language processing technology. The semantic analysis of the regulation instruction needs to design a structural instruction template, the traditional method is mainly to design the structural template based on the ground-air conversation rule, and extract keywords such as basic regulation terms, subjects, predicates and the like in the instruction by a keyword extraction method, the method cannot describe the association among different component words, and complex regulation instructions containing a plurality of verbs with different regulation intentions cannot be understood more accurately, so that the semantic analysis and understanding process of the actual regulation instruction is limited.
Disclosure of Invention
The purpose of the invention is as follows: the method analyzes the structure of the actual control instruction from the perspective of linguistics, extracts the control instruction by combining the land-air conversation rule and the natural language processing technology to form a computer-understandable structural instruction, and focuses on carrying out fine semantic role marking on the multi-verb control instruction. The control instructions generated under the actual working environment can be mostly analyzed.
The technical scheme is as follows: the invention classifies and extracts a plurality of different regulatory intentions and corresponding argument words contained in the complex regulatory instruction by using the fully-connected neural network, and understands the different regulatory intentions and the corresponding argument words through the semantic network, thereby being helpful for a computer to understand the intention of the complex regulatory instruction. The invention can be used for auxiliary alarm for scene conflicts caused by error, forgetting and omission of controllers. Especially for understanding of complex policing instructions with multiple policing intent verbs, and combining and converting the policing intents into logical expression form triples that can be understood by the system, such as policing instructions: CES3984, down to 13000 meters for maintenance. The control instruction comprises two different control intentions, namely descending and keeping, other words and words have different relations with the two words, and the neural network is required to classify the words in the control instruction, which are respectively related to the different control intentions. When the invention is used for the application of auxiliary alarm for scene conflicts caused by error, forgetting and omission of controllers, the invention comprises the following steps:
step 1, carrying out voice recognition processing on the control voice to obtain a control instruction in a text format;
step 2, performing part-of-speech analysis on words contained in the control instruction in the text format;
step 3, extracting a control intention based on the part of speech analysis result, and simultaneously extracting other words in the control instruction to form a candidate argument set;
step 4, carrying out BIO labeling processing on the candidate argument set (BIO labeling: labeling each element as 'B-X', 'I-X' or 'O', wherein 'B-X' indicates that the segment where the element is located belongs to the X type and the element is at the beginning of the segment, 'I-X' indicates that the segment where the element is located belongs to the X type and the element is at the middle position of the segment, 'O' indicates that the element does not belong to any type), adding E label to indicate that the argument has semantic relation with a plurality of verbs in the sentence at the same time, training parameters of a fully-connected neural network by using labeled data, and obtaining a control intention control verb-element group through the neural network;
step 5, judging the number of verb regulatory intentions in the regulatory instruction, and entering step 6 if the number is more than 1;
step 6, generating an input word vector;
step 7, classifying the input word vectors by using a neural network to obtain a probability output value z;
step 8, judging the relation: if the probability output value z has higher probability in a certain category, judging that the argument has a relationship with the corresponding control intention, executing the step 10, otherwise, not processing;
step 9, because the semantic relation between the control intention verb and different argument words is univocal mapping in most cases in the control instruction, a semantic network is constructed according to experience, and the semantic network comprises the control intention verb, the argument words and semantic association triple knowledge;
step 10, substituting the verb of the control intention and the argument word into a semantic network to obtain a specific semantic relation between the verb and a corresponding argument and form a structured control instruction;
and step 11, using the structured control instruction to directly detect whether scene conflict is caused by wrong, forgotten and missed issuing of wrong control instruction by a controller. For example: after the structured control instruction is obtained, the running track of the aircraft can be obtained by utilizing the information related to the words, whether scene conflict can occur or not can be judged by comparing whether the tracks obtained from the two control instructions are intersected, and if the control instruction comprises a plurality of control intents, the running track of the aircraft can be more accurately described by combining the control intents.
In the invention, the step 2 comprises the following steps:
and 2-1, performing Chinese word segmentation operation on the control instruction by utilizing the jieba word segmentation to obtain a word sequence.
Step 2-2, part of speech tagging: and labeling each word in the word sequence according to the corresponding part of speech to obtain a part of speech analysis result. The main parts of speech include: basic control terms (sp), verbs (v/vi), other parts of speech and the like, wherein in the step, special parts of speech are set due to the particularity of the control instructions, so that different words in the computer mechanism control instructions are helped to be solved;
the step 3 comprises the following steps: based on the result of part-of-speech analysis, removing words which are basic control terms in the control instruction, wherein one control instruction after the basic control terms are removed comprises verbs and part-of-speech words of other components, judging according to the result of part-of-speech tagging, extracting the verbs to form a verb set, and simultaneously extracting the other words to form a candidate argument set, wherein elements contained in the verb set and the candidate argument set are used for judging whether correlation exists or not, namely verb-argument correlation pairs;
judging which words in the instruction are basic control terms through the part of speech, if the words with the part of speech of "sp" are the basic control terms, extracting the basic control terms, and some basic control terms need to form a phrase structure with surrounding words, such as: the ground wind is 300 meters per second, special part-of-speech tagging is carried out on the control terms, a phrase searching algorithm is designed to obtain a control term phrase, the behavior sp of the ground wind is set by taking the control terms as an example, meanwhile, a backward searching rule is designed, and finally the ground wind is 300 meters per second and extracted as a phrase;
and judging according to the part-of-speech tagging result, extracting verbs to form a verb set, and extracting words of other components to form a candidate argument set. Removing light verbs from the extracted verbs, forming linkage structural words and the like, and storing results into a verb set;
step 4 comprises the following steps:
step 4-1, converting into word vectors: the word is preprocessed by using a word2vec method, words in a word vector expression control instruction are generated, word vector representation of an input sentence is generated and is used as input of a neural network, and due to the defects that high-dimensional and sparse one-hot representation (one-hot representation) word vectors cannot well represent word similarity and dimension disasters are easily caused, low-dimensional and dense distributed expression (distributed representation) word vectors are used for representing words. A commonly used distributed word vector is a 100-dimensional array of numbers.
Step 4-2, the invention mainly uses the fully-connected neural network, trains the model parameter of the fully-connected neural network: collecting control instructions in an actual working scene and forming a corpus training set, manually labeling the corpus training set of the control instructions, and training weight parameters of neurons in all layers of the fully-connected neural network by using the corpus training set to obtain a trained neural network; when marking data, marking different categories of the governing intention verbs, such as B0, B1, … and BM, wherein M is the category of the governing intention, such as: CES3421, down to 5000 meters retention. Two regulatory intention verbs "drop" and "hold" are included, the verb "drop" is labeled B0, the verb "hold" is labeled B1, and the argument words associated with the verb are labeled the same numeric category, such as "to" is labeled B0;
and 4-3, predicting by using a neural network: judging whether the target argument in the candidate argument set and the verb set is associated with the verb through the trained neural network to obtain a verb-argument, for example: at regulatory command "CES 3984, runway 35L, take off immediately. "in, the governing-intention verb is takeoff, and the arguments include" CES3984, runway 35L ", and the verb-argument group here is respectively: takeoff-CES 3984 and takeoff-runway 35L.
In step 5, the number of verbs in the control command is determined, and the verb in the step does not include the light verb, and the interlocked structural word is regarded as one verb. If the number of verbs is 1, the detailed subsequent processing refers to a patent 'a structured regulation instruction extraction method based on natural language processing', the invention aims at the condition of a plurality of regulation intention verbs, and if the number of verbs is more than 1, the step 6 is entered;
the step 6 comprises the following steps: converting the input sentence into word vector input;
in step 7, the trained neural network is used to classify the input matrix. The neural network can capture the internal features of the input word vector well. Because the judgment of whether the target verb and the target argument have a relationship can be regarded as a two-classification problem, the output end of the neural network contains probability output values corresponding to different classes of regulatory intents: if the probability is higher, the target argument is in relation with the selected target verb, otherwise, the target argument is in relation with the selected target verb;
the neural network comprises an input layer, a hidden layer and an output layer, and is defined to have 100 × N input x, wherein N represents the length of a sentence, and a maximum length N is generally defined to be 30, that is, the maximum length contains 30 words; if there are n neurons in the input layer, defining the weight matrix as W(2w+2)×nBias matrix is b1×nThe output value of the input layer is alpha1×nThe output value of the activation function is h1×nDefining a ReLU activation function fRe Lu(t) is:
where t is the input value, the output value of the neuron of the hidden layer is obtained by the following formula:
αh＝hhWh+bh，
hh＝fRe Lu(αh)，
wherein the weight matrix of the hidden layer is WhBias matrix is bhOutput value of alphahThe output value of the activation function is hhThe output value z of the neural network is obtained by:
z＝hhWo+bo，
wherein the weight matrix of the output layer is WoBias matrix is boThe output value is z.
Step 9 comprises: the construction idea of The Semantic net of The control instructions is derived from frame semantics (cited documents: Grigoris.A, Paul.G, Frank.v.H, Rinke.H.A Semantic Web Primer (Third Edition) [ M ]. Cambridge: The MIT Press, 2012: 1-288.), and The basic theory is The lattice syntax theory of The American linguistics' Fillmore. The definition of the verb grid in the semantic network can be used for more finely describing the relationship between the argument and the verb in the regulation instruction, and generally comprises the following steps: a stop grid, a source grid, a place grid, an end grid, etc. Because the semantic relation between the control intention verb and different argument words is unilaterally mapped in most cases in the control instruction, a semantic network is constructed according to experience, and the semantic network comprises the control intention verb, the argument words and semantic association triple knowledge;
the invention also comprises a step 12 of constructing a structured template:
the structured template herein comprises two parts: 1. a basic governing term phrase; 2. verb-argument-lattice triplets. Because the control instruction comprises a plurality of verbs, the second part is composed of verb-argument relation triples according to the appearance sequence of the verbs. And establishing a structured template with two parts of basic control terms and verb-argument-lattice triples, and filling deconstructed control instruction information into the template to form a structured instruction.
The method can be applied to semantic understanding of the control command in the air traffic control system. Since a plurality of verbs may be included in the regulation instruction issued by the regulator in actual work, judging arguments associated with different verbs and finding out the relationship between them are important for accurately understanding the regulation instruction. The method can better analyze the control command and form a structured command, and can effectively process the work from the speech recognition of the control command to the movement trend prediction based on the control command content.
Has the advantages that: the invention has the following technical effects:
1. and (4) enabling the computer to autonomously understand the semantics of the control instruction and judge the motion process of the aircraft.
2. Semantic role labeling is performed on a plurality of verbs appearing in the management instruction.
3. Basic regulation term information appearing in the regulation instruction is extracted.
4. And converting the unstructured regulatory instruction into a structured regulatory instruction.
Drawings
The foregoing and other advantages of the invention will become more apparent from the following detailed description of the invention when taken in conjunction with the accompanying drawings.
FIG. 1 is a neural network-based regulatory instruction multi-verb role classification and structured instruction extraction process.
FIG. 2 is a main flowchart of a method for extracting a structured policing instruction.
Fig. 3 is a structure of a neural network.
FIG. 4 is a generated structured policing instruction template.
Detailed Description
The invention is further explained below with reference to the drawings and the embodiments.
The present invention is described in further detail below with reference to the use cases of the policing instructions and the associated figures. First, an example of a policing instruction is given: CES1234, eastern tower, slides along taxiways d5-p4-a5 to the 18 left runway waiting point, waiting outside the 18 left runway.
For ease of illustration and description, the steps of implementation herein are divided according to the main flow diagrams shown in fig. 1 and 2.
Step 1: part of speech analysis
The method comprises two steps of treatment processes: chinese word segmentation and part-of-speech tagging. And carrying out Chinese word segmentation and part-of-speech tagging on the control instruction by using a jieba word segmentation in Python software to obtain a corresponding result. In the word segmentation process, basic control terms are added in the word segmentation dictionary, so that the precision of the word segmentation result can be improved.
Step 2: extraction of words from each component
In the last step, words of different composition are labeled as different parts of speech, wherein the part of speech of "east tower" is labeled as "sp", "slide to" and "wait" are labeled as "v". According to the part of speech, the oriental tower is a basic control term and is extracted. The original regulatory terms become: CES1234, sliding along taxiway d5-p4-a5 to the 18 left runway waiting point, waiting outside the 18 left runway, closer to natural language. Get verbs "slide to" and "wait" at the same time. Other component words in the sentence are extracted as candidate arguments.
And step 3: relationship determination
The step is taken as the core step of the invention, and is mainly introduced in three parts:
and 3-1, converting the word vector into a word vector to obtain an input vector. According to the principle of word2vec, words can train out corresponding low-dimensional and dense distributed vectors through a neural network, and the relation among similar words can be obtained through the distribution of word vectors. Word vectors for each word can be calculated from regulatory instruction text data using the gensim framework package in Python software. It should be noted that the amount of data in the text data set of the regulatory directive is large enough to contain all of the regulatory directive vocabulary (not including the basic regulatory terms).
And 3-2, training model parameters of the neural network. The corpus training set of the control instruction is labeled, and the input vector not only considers the relation between the target argument and the target verb in the sentence, but also considers the function of adjacent words of the target argument, namely the function of context information. If the target argument is associated with the target verb when in a context, the tag is defined as the same tag as the target argument. It should also be noted that the number of positive and negative samples should be equal when preparing the training set. The fully-connected neural network was written using Python, the structure is shown in fig. 3, and multiple rounds of training were performed on the parameters using the training data. The larger the amount of training data, the better the model.
And 3-3, predicting by using a neural network. In this step, two models can be used: one is a word vector model which is responsible for converting words into word vectors, and the other is a fully-connected neural network classification model which is responsible for judging whether the target argument is associated with the verb. Through this step, verb-theoretic tuples can be obtained, but at this time the relationship between the two is not known.
And 4, step 4: semantic Web construction
The step completes the relation between the associated verb and the argument by constructing a semantic network. The semantic web is constructed in the form of an ontology and can express the relationship between the verb and other entities in the control field. The lattice types of verbs in the present invention include: the site lattice, the object lattice, the place lattice, the direction lattice, the source point lattice and the end point lattice. Since the field of empty management belongs to a special field, the format type carried by the verb appearing in the regulatory instruction is not the same as that in the normal natural language environment, such as: and (4) sliding. Two example sentences are given below:
1. CES1234, sliding to the end of the runway.
2. He steps on the skateboard to glide.
In the first sentence, the regulation instruction shows that the 'slide' in the regulation instruction has no tool lattice, and in the second sentence, the 'slide' is the tool lattice of the verb 'slide'.
Because the definition between verbs and arguments in the semantic web is fixed, and the numbers such as flight number, height, speed and the like in the control instruction do not have fixed values, another advantage of using word vectors to express words is that the similarity of similar words can be expressed, and the problem can be solved by setting templates and using the similarity of word vectors.
And 5: structured instructions
Through the processing of the previous steps, the deconstructed policing instruction is obtained, wherein the verb is judged to have two, so the structured policing instruction in the example is divided into three parts:
the basic regulatory terms: an east tower.
Verb 1: slide to CES1234, schlieren, slide to d5-p4-a5, slide to wait point to end point grid.
Verb 2: wait-CES 1234-shit grid, wait-off-runway-azimuth grid.
The resulting structured template results are shown in fig. 4.
From the results, flight CES1234 performs a taxi action first, and the taxi action is performed on taxiway d5-p4-a5, and its taxi end point is a runway waiting point, and performs a second action: and waiting for action. And obtaining a sliding path and an end point of flight CES1234 according to the analysis.
Examples
For convenience of illustration and description, the steps implemented herein are divided according to the main flow chart shown in fig. 2, and are explained in conjunction with the actual policing instructions. First, an example of a policing instruction is given:
1. CQH1207, east tower, sliding along D5-P4, waiting outside runway 35L.
2. CES3984, east tower, runway 35L, may take off.
Step 1, carrying out voice recognition processing on the control voice to obtain a control instruction in a text format; the policing voice is processed using a speech recognition device resulting in unstructured policing instruction text, as shown above.
Step 2, performing part-of-speech analysis on words contained in the control instruction in the text format, for example:
1. CQH/eng,1207/m, east tower/sp, edge/P, D5-P4/m, glide/v, runway/n, 35L/m, outer/f, wait/v.
2. CES/eng,3984/m, east tower/sp, runway/n, 35L/m, and take-off/v.
The part-of-speech tagging is carried out on the words in the control instruction, so that the part-of-speech of different words can be obtained, the control-purpose words and other words can be distinguished, meanwhile, the different words are combined into word groups in a mode of using a well-defined control-instruction special word group dictionary, and if: CQH1207, CES3984, runway 35L.
And 3, extracting a control intention based on the part-of-speech analysis result, and simultaneously extracting other words in the control instruction to form a candidate argument set, wherein the word of the verb part-of-speech is usually the control intention, for example: taxiing, waiting and taking off. Other words, such as: CQH1207, CES3984, east tower, runway 35L, edge, D5-P4, outer, etc. are all argument words.
Step 4, performing BIO labeling processing on the candidate argument set, and training parameters of the fully-connected neural network by using labeled data, wherein the goal is to obtain a control intention verb-argument group through the neural network, for example: the control instruction 1 includes two verbs, which are denoted as X ═ 0 and X ═ 1, the CQH1207 is associated with both verbs, and the control instruction E is denoted, the east tower is not associated with both verbs, and the control instruction O is denoted, and the control instruction B-0 is denoted, along with the association with only the glide.
Step 5, judging the number of verb regulatory intentions in the regulatory instruction, and entering step 6 if the number is more than 1;
and 6, generating an input word vector, obtaining the number sequence word vectors of different words by using word2vec, and forming the word vectors of sentences by using the word vectors of the words.
Step 7, classifying the input word vectors by using a neural network to obtain a probability output value z;
step 8, judging the relation: if the probability output value z has a higher probability in a category, it is determined that the argument has a relationship with the corresponding regulatory intention, step 10 is executed, otherwise, no processing is performed, for example: if the edge is associated with glide only and not with wait in step 4, then its probability at marker B-0 will be much greater than at marker B-1.
Step 9, constructing a semantic network, wherein the semantic network comprises regulation intention verbs, argument words and semantic association triple knowledge, such as: the semantic relationship between the flight and the verb is the operator of the action, the relationship between the runway 35L and the takeoff capability is the starting position of the action, and the relationship between D5-P4 and the taxi is the path of the action implementation.
Step 10, substituting the verb of the control intention and the argument word into a semantic network to obtain a specific semantic relation between the verb and a corresponding argument and form a structured control instruction;
step 11, the structured regulation instruction is used to directly detect whether a scene conflict is caused by the wrong, forgotten, or missed sending of the wrong regulation instruction by the controller, for example: when the controller sends the above two instructions in sequence, the following structured instruction can be obtained from instruction 1:
CQH1207, glide: the implementer;
CQH1207, wait for: the implementer;
D5-P4, slide: an action path;
outside runway 35L, wait: an action location;
then the end of the taxi is outside the runway 35L, as can be seen from the meaning of waiting.
From instruction 2, the following structured instructions can be derived:
CES3984, can take off: the implementer;
runway 35L, which can take off: an action source point;
therefore, the paths represented by the two instructions can be obtained, and the two control instructions are judged not to conflict.
The method and the device can be used for analyzing the intention of the empty control command, and the control command containing a plurality of different verbs can be processed by using the method and the device to obtain the structured commands with the same number as the verbs. The invention can be used for auxiliary alarm for scene conflicts caused by error, forgetting and omission of controllers. The semantic lattice analyzes the action intention of the control command and the action route of the aircraft, so that whether the conflict between the aircraft can be caused by different control commands can be judged.
The present invention provides a classification method for regulatory instructions based on semantic network, and a plurality of methods and approaches for implementing the technical scheme, where the foregoing is only a preferred embodiment of the present invention, it should be noted that, for those skilled in the art, a plurality of improvements and modifications may be made without departing from the principle of the present invention, and these improvements and modifications should also be regarded as the protection scope of the present invention. All the components not specified in the present embodiment can be realized by the prior art.
Claims (1)
1. A method for classifying control instructions based on a semantic network is characterized by comprising the following steps:
step 1, carrying out voice recognition processing on the control voice to obtain a control instruction in a text format;
step 2, performing part-of-speech analysis on words contained in the control instruction in the text format;
step 3, extracting a control intention based on the part of speech analysis result, and simultaneously extracting other words in the control instruction to form a candidate argument set;
step 4, carrying out BIO labeling processing on the candidate argument set, and training parameters of a fully-connected neural network by using labeled data, wherein the goal is to obtain a control intention verb-argument set through the neural network;
step 5, judging the number of verb regulatory intentions in the regulatory instruction, and entering step 6 if the number is more than 1;
step 6, generating an input word vector;
step 7, classifying the input word vectors by using a neural network to obtain a probability output value z;
step 8, judging the relation: if the probability output value z has higher probability in a category, judging that the argument has a relationship with the corresponding control intention, executing the step 10, otherwise, not processing;
step 9, constructing a semantic network, wherein the semantic network comprises a regulation intention verb, argument words and semantic association triple knowledge;
step 10, substituting the verb of the control intention and the argument word into a semantic network to obtain a specific semantic relation between the verb and a corresponding argument and form a structured control instruction;
step 11, the structured control instruction is used for directly detecting whether scene conflict is caused by wrong, forgotten and missed issuing of wrong control instructions by a controller;
the step 2 comprises the following steps:
step 2-1, performing Chinese word segmentation operation on the control instruction by utilizing jieba word segmentation to obtain a word sequence;
step 2-2, part of speech tagging: labeling each word in the word sequence according to the corresponding part of speech to obtain a part of speech analysis result;
the step 3 comprises the following steps: based on the result of part-of-speech analysis, removing words which are basic control terms in the control instruction, judging a control instruction which contains verbs and words of parts-of-speech of other components after the basic control terms are removed, extracting the verbs to form a control intention verb set, and simultaneously extracting the words of other components to form a candidate argument set;
in step 4, the BIO labeling means: labeling each element as B-X, I-X or O, wherein B-X indicates that the fragment in which the element is located belongs to X type and the element is at the beginning of the fragment, I-X indicates that the fragment in which the element is located belongs to X type and the element is in the middle of the fragment, and O indicates that the element does not belong to any type;
step 4 comprises the following steps:
step 4-1, converting into word vectors: preprocessing the words by using a word2vec method, and generating word vector representation of an input sentence as input of a neural network;
step 4-2, using the fully-connected neural network to train model parameters of the fully-connected neural network: collecting control instructions in an actual working scene and forming a corpus training set, labeling the corpus training set of the control instructions, and training weight parameters of neurons of all layers of the fully-connected neural network by using the corpus training set to obtain a trained neural network;
and 4-3, predicting by using a neural network: judging whether the target argument in the candidate argument set and the verb set is associated with the verb through the trained neural network to obtain a verb-argument;
the step 6 comprises the following steps: using the sentence word vector obtained in the step 4-1 as the input of the neural network model
In step 7, the neural network comprises an input layer, a hidden layer and an output layer, the neural network is defined to have 100 x n input x, if the input layer has n neurons, the weight matrix is defined as W(2w+2)×nBias matrix is b1×nThe output value of the input layer is alpha1×nThe output value of the activation function is h1×nDefining a ReLU activation function fReLU(t) is:
where t is the input value, the output value of the neuron of the hidden layer is obtained by the following formula:
αh＝hhWh+bh，
hh＝fReLU(αh)，
wherein the weight matrix of the hidden layer is WhBias matrix is bhOutput value of alphahThe output value of the activation function is hhThe output value z of the neural network is obtained by:
z＝hhWo+bo，
wherein the weight matrix of the output layer is WoBias matrix is bo。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910348919.2A CN110232121B (en) | 2019-04-28 | 2019-04-28 | Semantic network-based control instruction classification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910348919.2A CN110232121B (en) | 2019-04-28 | 2019-04-28 | Semantic network-based control instruction classification method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110232121A CN110232121A (en) | 2019-09-13 |
| CN110232121B true CN110232121B (en) | 2021-04-06 |
Family
ID=67860304
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910348919.2A Active CN110232121B (en) | 2019-04-28 | 2019-04-28 | Semantic network-based control instruction classification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110232121B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110837731A (en) * | 2019-10-12 | 2020-02-25 | 创新工场(广州)人工智能研究有限公司 | Word vector training method and device |
| CN111179926A (en) * | 2019-12-10 | 2020-05-19 | 深圳微品致远信息科技有限公司 | Method and device for generating aircraft control command and computer equipment |
| CN113569545B (en) * | 2021-09-26 | 2021-12-07 | 中国电子科技集团公司第二十八研究所 | Control information extraction method based on voice recognition error correction model |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5987404A (en) * | 1996-01-29 | 1999-11-16 | International Business Machines Corporation | Statistical natural language understanding using hidden clumpings |
| CN107656990A (en) * | 2017-09-14 | 2018-02-02 | 中山大学 | A kind of file classification method based on two aspect characteristic informations of word and word |
| CN109460547A (en) * | 2018-09-19 | 2019-03-12 | 中国电子科技集团公司第二十八研究所 | A kind of structuring control order extracting method based on natural language processing |
-
2019
- 2019-04-28 CN CN201910348919.2A patent/CN110232121B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5987404A (en) * | 1996-01-29 | 1999-11-16 | International Business Machines Corporation | Statistical natural language understanding using hidden clumpings |
| CN107656990A (en) * | 2017-09-14 | 2018-02-02 | 中山大学 | A kind of file classification method based on two aspect characteristic informations of word and word |
| CN109460547A (en) * | 2018-09-19 | 2019-03-12 | 中国电子科技集团公司第二十八研究所 | A kind of structuring control order extracting method based on natural language processing |
Non-Patent Citations (2)
| Title |
|---|
| 卢薇冰.基于CNN的陆空通话语义识别方法.《中国优秀硕士学位论文全文数据库》.2017,(第1期),第C031-283页. * |
| 王煊.用于管制语音理解的语义分析方法.《指挥信息系统与技术》.2019,第10卷(第1期),第33-36页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110232121A (en) | 2019-09-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Lin et al. | A real-time ATC safety monitoring framework using a deep learning approach | |
| Chen et al. | Research on text sentiment analysis based on CNNs and SVM | |
| CN110232121B (en) | Semantic network-based control instruction classification method | |
| CN110287494A (en) | A method of the short text Similarity matching based on deep learning BERT algorithm | |
| CN110232114A (en) | Sentence intension recognizing method, device and computer readable storage medium | |
| CN109376242A (en) | Text classification algorithm based on Recognition with Recurrent Neural Network variant and convolutional neural networks | |
| CN109684626A (en) | Method for recognizing semantics, model, storage medium and device | |
| CN107818153A (en) | Data classification method and device | |
| CN110188345B (en) | Intelligent identification method and device for electric operation ticket | |
| CN113157859B (en) | Event detection method based on upper concept information | |
| CN111475650B (en) | Russian semantic role labeling method, system, device and storage medium | |
| CN113569545B (en) | Control information extraction method based on voice recognition error correction model | |
| CN108536781B (en) | Social network emotion focus mining method and system | |
| Abedin et al. | Cause identification from aviation safety incident reports via weakly supervised semantic lexicon construction | |
| Luo et al. | Using lda2vec topic modeling to identify latent topics in aviation safety reports | |
| CN110321434A (en) | A kind of file classification method based on word sense disambiguation convolutional neural networks | |
| CN110069771B (en) | Control instruction information processing method based on semantic chunk | |
| Jing et al. | BERT for aviation text classification | |
| CN113948091A (en) | Air-ground communication voice recognition engine for civil aviation passenger plane and application method thereof | |
| CN109460547B (en) | Structured control instruction extraction method based on natural language processing | |
| CN113642862A (en) | Method and system for identifying named entities of power grid dispatching instructions based on BERT-MBIGRU-CRF model | |
| CN113158658B (en) | Knowledge embedding-based structured control instruction extraction method | |
| CN109344251A (en) | A kind of particular text information extraction method based on layer classifier and template matching | |
| CN112200268A (en) | Image description method based on encoder-decoder framework | |
| Mesmia et al. | Transducers cascades for an automatic recognition of Arabic named entities in order to establish links to free resources |
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 |