A method for constructing a patent technology classification system in the field of air traffic management

Abstract

The present invention discloses a method for constructing a patent technology classification system in the field of air traffic management, and belongs to the field of patent analysis and mining technology. The method comprises step 1, mining key technology classification vocabulary in the field of air traffic management; step 2, calculating the measurement value of the technology classification vocabulary; step 3, constructing an initial technology classification system for patents in the field of air traffic management; step 4, constructing an expert set, and the experts in the expert set evaluate the technology classification vocabulary to obtain a first evaluation matrix; step 5, reducing the expert set and calculating a third evaluation matrix; step 6, calculating the importance and consistency of the technology classification vocabulary; step 7, reducing the technology classification vocabulary set to obtain the final patent technology classification system. The method makes full use of the domain corpus and the knowledge accumulation of experts, and reduces the bias of expert decision-making, so that the patent technology classification system finally constructed is more representative.

Description

Technical Field

[0001] The present invention belongs to the field of patent analysis and mining technology, and in particular, relates to a method for constructing a patent technology classification system in the field of air traffic management. Background Art

[0002] As an important part of the national air defense system and the air transportation system, air traffic management is the command center for ensuring flight safety, implementing air traffic management and services, and realizing efficient and orderly operation of air transportation. Its performance represents the technical level of the air transportation industry. As air flights become increasingly complex, flight conflicts, large-scale airspace congestion and other problems occur frequently. Therefore, there is an urgent need for an air traffic management system with core key technologies with independent intellectual property rights and independent research and production capabilities.

[0003] The first prerequisite for designing and laying out a high-value patent system in the field of air traffic management is to conduct a comprehensive and detailed search and analysis of patents in the field of air traffic management. In the current international patent classification, patents related to air traffic management are scattered under the classification numbers of multiple groups, not concentrated in one large group, and the classification granularity is not fine enough. Therefore, the international patent technology classification system cannot be used as an input for patent search analysis. In addition, keyword retrieval is mainly based on the personal experience of experts and is easily affected by personal cognitive level and preferences. Therefore, it is urgent to establish a patent technology classification system in the field of air traffic management to support the analysis and layout of patents. Existing Technology 1 (Zhu Yongwen, Chen Zhijie, et al. Research on Scientific and Technological Issues of Intelligent Air Traffic Management [J]. China Engineering Science, 2023, 25(5): 174-184.) analyzes the challenges faced by traditional air traffic management from the three directions of traffic control, airspace management, and flow control, summarizes the basic scientific issues involved in intelligent air traffic management, and suggests integrating and applying cutting-edge technologies such as satellite Internet, big data, digital twins, and cloud computing to build an intelligent air traffic management technology system from the aspects of aircraft, airspace, control decision-making, and operation, thereby establishing the technical foundation of a new generation of air traffic management systems. Existing Technology 2 (Zhao Binbo. Analysis of Key Technologies and Several Issues of Collaborative Air Traffic Flow Management [J]. Electronic World, 2019, (20): 201-202.) proposes that the key technologies of collaborative air traffic flow management include collaborative ATFM technology application, collaborative ground holding technology, collaborative airport surface management strategy, collaborative route change strategy, and collaborative approach and departure capacity optimization model. Existing Technology 3 (Deng Qiwen. Research on Key Technologies of Airspace Planning in Air Traffic Management [J]. China New Communications, 2018, 20(3): 47-48.) studied airspace planning technologies, including airspace operation safety assessment technology, airspace operation characteristic analysis technology, route network design planning and real-time optimization technology, communication navigation surveillance and support technology, etc. Existing Technology 4 (Tian Yong. Research on Key Technologies of Air Traffic Flow Management [D]. Nanjing: Nanjing University of Aeronautics and Astronautics, 2009.) studied air traffic flow management, including air traffic capacity assessment, ground holding, rerouting strategy, terminal area sorting and other technologies.

[0004] The above research results classified and studied the sub-fields of air traffic management from different perspectives. The classification of technical systems is mainly based on the personal experience of experts, so the relevant results are easily affected by personal cognitive levels and preferences. Summary of the invention

[0005] Purpose of the invention: The technical problem to be solved by the present invention is to provide a method for constructing a patent technology classification system in the field of air traffic management in view of the shortcomings of the prior art.

[0006] In order to solve the above technical problems, the present invention discloses a method for constructing a patent technology classification system in the field of air traffic management, comprising the following steps:

[0007] Step 1: mining key technical classification vocabulary in the field of air traffic management to obtain a technical classification vocabulary set;

[0008] Step 2, calculating the measurement value of the technical classification vocabulary in the technical classification vocabulary set;

[0009] Step 3: Construct an initial technical classification system for patents in the field of air traffic management based on the measurement values ​​of technical classification vocabulary;

[0010] Step 4, constructing an expert set, wherein the experts in the expert set evaluate the technical classification vocabulary to obtain a first evaluation matrix;

[0011] Step 5, reducing the expert set according to the first evaluation matrix to calculate a third evaluation matrix;

[0012] Step 6, calculating the importance and consistency of the technical classification vocabulary according to the third evaluation matrix;

[0013] Step 7, reducing the technical classification vocabulary set according to the importance and consistency of the technical classification vocabulary to obtain the final patent technology classification system.

[0014] Furthermore, step 1 includes: constructing a corpus in the field of air traffic management D ={ d 1, d 2,…, d k ,…, d K}, K For the corpus D Total number of documents in d k Representation Corpus D Middle k documents, 1 ≤ k ≤ K ; From the corpus D Extract keywords from the data and process them to form a technical classification vocabulary set in the field of air traffic management W ={ w 1, w 2,…, w i ,…, w N}, N is the total number of technical classification words, w i Vocabulary of technical classification W The i-th word in , 1 ≤ i ≤N .

[0015] Furthermore, the measurement value of the technical classification vocabulary in step 2 includes the co-occurrence frequency of two words in the technical classification vocabulary set and the importance of the words; the co-occurrence frequency of the two words is calculated Co ( w i , w j )The formula is as follows:

[0016] ,

[0017] in, w i , w j represents the vocabulary in the technical classification vocabulary set W, 1 ≤ i , j ≤ N ; When the vocabulary w i and w j Also appears in the document d k middle, P k ( w i , w j )=1, otherwise P k ( w i , w j )=0; Calculate vocabulary w i Importance g ( w i )The formula is as follows:

[0018] .

[0019] By calculating the co-occurrence frequency and importance of words, the potential semantic relationship between words is mined from the corpus. The higher the co-occurrence frequency, the more likely there is a direct relationship between the semantic concepts of the two words; the higher the importance, the more likely the semantic concept of the word is a top-level concept. This provides a basis for building parent-child relationships between words.

[0020] Furthermore, step 3 includes constructing the parent-child relationship of vocabulary to form an initial technical classification system: for the technical classification vocabulary set W Any word in w c , for which the parent node vocabulary is selected wp , if and only if ; In each child node w c and the parent node w p Add an edge between e cp , forming the initial technology classification system T 0={ W , E},in, . By constructing the parent-child relationship of vocabulary, a tree structure of vocabulary is formed.

[0021] By constructing a tree structure of vocabulary, the domain knowledge contained in the hierarchical technology classification system is enriched. The closer to the root of the tree, the more top-level the semantic concept of the vocabulary; the parent-child relationship of the nodes in the tree expresses the inclusion relationship of the semantic concepts of the vocabulary. This provides more complete and informative input for the subsequent subjective evaluation of domain experts.

[0022] Furthermore, step 4 includes forming an expert set S ={ s 1, s 2,…, s m ,…, s M Initial Technology Classification System T 0 for evaluation, where s m Indicates m Experts, m =1,2,…, M During the evaluation process, ensure M The evaluation experts are independent of each other and construct the following first evaluation matrix V :

[0023] ;

[0024] in, For the m Experts s m For i Technical Classification Vocabulary w i Weighted assessment of importance.

[0025] Furthermore, in step 4 m Experts s m For i Technical Classification Vocabulary w i Weighted Assessment of Importance Taking into account the expert ratings and w i Importance g ( w i ), the calculation formula is as follows:

[0026]

[0027] For the m The expert has the i-th technical classification vocabulary w i The evaluation of importance ranges from {unimportant, moderately important, important, very important, very important}, and the corresponding values ​​are 1, 3, 5, 7, 9; g’ ( w i ) is the i Technical Classification Vocabulary w i The normalized importance of ; or is the weight coefficient, which is the weight of controlling the importance of the vocabulary. Calculate the weighted assessment Considering the expert ratings and w i Importance g ( w i ), reducing the influence of experts’ personal cognitive level and preferences.

[0028] Further, step 5 includes: judging whether the deviation of the comprehensive evaluation results of all experts is greater than a set threshold, m Deviation of the evaluation results of experts d m The calculation formula is:

[0029]

[0030] In the formula,

[0031] d m Indicates m Experts' familiarity with current assessment content, d m The larger the m The less familiar the experts are with the current assessment content, the lower the threshold of expert concentration reduction is. d 0∈(0,1]; if d j > d 0, the expert is deemed not familiar with the current assessment content and is removed; otherwise, he is retained.

[0032] Thus, the reduced expert set is obtained as , 1 ≤ r ≤ M By eliminating experts who are not familiar with the current evaluation content, the experts' weighted assessment is made more accurate, thus ensuring that the final technology classification system is more representative.

[0033] Furthermore, step 5 also includes: removing the row vectors corresponding to the removed experts in the first evaluation matrix V, and obtaining the second evaluation matrix as :

[0034]

[0035] After normalization, the third evaluation matrix is ​​obtained for:

[0036]

[0037] in: , i =1,2,…, N , m =1,2,…, r , v max The second evaluation matrix The maximum value of the weighted evaluation values ​​of the importance of technical classification terms by experts in v min It is the minimum value among the weighted evaluation values ​​of the importance of technical classification vocabulary by experts.

[0038] Further, step 6 includes: according to the third evaluation matrix Comprehensive calculation of each technical classification vocabulary w i Importance v i :

[0039]

[0040] The standard deviation is used to measure each technical classification vocabulary w i The consistency of the corresponding multiple evaluation results s i :

[0041] .

[0042] Further, step 7 includes: classifying the vocabulary according to the technology w i Importance v iand consistency s i Air Traffic Management Technology Classification Set W ={ w 1, w 2,…, w N} to reduce and set the importance threshold of technical classification vocabulary v 0∈(0,1) and the consistency threshold of the evaluation results s 0∈(0,1), when s i < s 0, v i < v When it is 0, the technical classification vocabulary is considered unimportant and is directly removed from the technical classification vocabulary set;

[0043] When satisfied s i < s 0, v i ≥ v When the value is 0, the technical classification vocabulary is considered important and should be retained in the technical classification vocabulary set:

[0044] when s i ≥ s 0, it indicates that the effectiveness of the evaluation result is poor and the choice of technical classification vocabulary cannot be determined. Return to step 4 for processing; if repeated execution If there are still controversial technical classification terms after this, they will be deleted to form the final technical classification vocabulary set. Parent-child relationship with vocabulary :

[0045]

[0046] Where 1 ≤ p ≤ N , Q The value of is 3~5;

[0047] Final patent technology classification system .

[0048] The evaluation matrix is ​​reconstructed using the reduced expert opinions, and the consistency of the evaluation opinions is calculated based on the standard deviation. Combined with the importance of the classification vocabulary, the selection of classification vocabulary is determined. The standard deviation, to a certain extent, represents the fluctuation of the expert's personal cognitive bias on the final result of the classification vocabulary. Therefore, through this iterative optimization based on the consistency of the standard deviation, the confidence level of the final result of the classification vocabulary can be improved.

[0049] Compared with the prior art, the present invention has the following significant advantages: the method of the present invention adopts a fusion method of "data-driven + expert knowledge". In the data-driven stage, the "implicit" knowledge of the domain corpus is mined, and in the expert comprehensive evaluation and optimization stage, the "explicit" knowledge of multiple experts is integrated, thereby ensuring that the final constructed technology classification system is more representative. Air traffic management includes multiple functions such as traffic control, airspace management, and flow management, involving knowledge in many fields. When constructing a patent technology classification system in the field of air traffic management, in order to reduce the influence of experts' personal cognitive level and preferences, the present invention uses multiple experts to construct a technology classification system through group decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments, and the above and / or other advantages of the present invention will become more clear.

[0051] Figure 1 A flow chart of a method for constructing a patent technology classification system in the field of air traffic management provided in an embodiment of the present application.

[0052] Figure 2 Another flowchart of a method for constructing a patent technology classification system in the field of air traffic management provided in an embodiment of the present application.

[0053] Figure 3 A schematic diagram of an initial technology classification system in a method for constructing a patent technology classification system in the field of air traffic management provided in an embodiment of the present application.

[0054] Figure 4 A schematic diagram of the final patent technology classification system in a method for constructing a patent technology classification system in the field of air traffic management provided in an embodiment of the present application. DETAILED DESCRIPTION

[0055] The embodiments of the present invention will be described below in conjunction with the accompanying drawings and specific implementation examples.

[0056] like Figure 1 As shown, the embodiment of the present application discloses a method for constructing a patent technology classification system in the field of air traffic management, such as Figure 2 As shown, it can be divided into two stages: data-driven and expert comprehensive evaluation optimization, including the following steps:

[0057] Phase 1: Data-driven generation of an initial technology classification system.

[0058] Step 1: Mining the key technical classification vocabulary in the field of air traffic management and obtaining a technical classification vocabulary set.

[0059] Building a corpus in the field of air traffic management D={ d 1, d 2,…, d k ,…, d K}, K For the corpus D Total number of documents in d k Representation Corpus D Middle k documents, 1 ≤ k ≤ K ; From the corpus D Extract keywords from the data and process them to form a technical classification vocabulary set in the field of air traffic management W ={ w 1, w 2,…, w i ,…, w N}, N is the total number of technical classification words, w i Vocabulary of technical classification W Middle i words, 1 ≤ i ≤ N In this embodiment, the domain corpus selects publicly published papers in the field of air traffic management, and uses the "abstract" of the paper as the corpus. D . The union of the paper's "keywords" is used as the technical classification vocabulary set W , thus eliminating interference such as stop words.

[0060] In the specific implementation process, the latest 500 papers were extracted from core journals in the field of "Civil Aviation Management" to form a field corpus , extract the keywords of the paper, and arrange them in reverse order according to the frequency of the words, and select the first 30 as the candidate air traffic management technology classification vocabulary set W ={air traffic control, air traffic control bureau, air traffic control center, airspace operation management, air traffic control system, air traffic control facilities, coordination mechanism, ground service, air traffic control safety, control operation, terminal area, airspace management, flow management, airport operation, air traffic control big data, control decision, airport digital twin, data processing, low-altitude control, airspace planning, airspace simulation evaluation, situation prediction, flow management method, flow management efficiency evaluation, taxiing, runway operation, airport digitization, data collection and cleaning, data analysis, data aggregation}. K The value is 500. N The value is 30.

[0061] Step 2: Calculate the metrics of the technical classification vocabulary in the technical classification vocabulary set.

[0062] The measurement values ​​of technical classification vocabulary mainly include the "co-occurrence frequency" of any two words in the technical classification vocabulary set and the "importance" of the words.

[0063] The formula for calculating the co-occurrence frequency of two words is ,in, w i , w j Vocabulary of technical classification W The vocabulary in , 1 ≤ i , j ≤ N ; When the vocabulary w i and w j Also appears in the document d k middle, P k ( w i , w j )=1, otherwise P k ( w i , w j )=0.

[0064] Computing vocabulary w i Importance g ( w i )The formula is as follows: .

[0065] Step 3: Construct an initial technical classification system for patents in the field of air traffic management based on the measurement values ​​of the technical classification vocabulary.

[0066] Based on the calculation results of the "co-occurrence frequency" of any two words in the technical classification vocabulary set and the "importance" of the words, the parent-child relationship of the words is constructed, thereby forming a tree structure of the words. , for which the parent node vocabulary is selected , if and only if , that is, select the importance ratio The node with the highest co-occurrence frequency is selected as its parent node In each child node and the parent node Add an edge between , forming the initial technical classification system ,in, In this embodiment, the initial technology classification system T 0 Figure 3 shown.

[0067] The second stage is the expert comprehensive evaluation and optimization stage.

[0068] Step 4: Construct an expert set. Experts in the expert set evaluate the technical classification vocabulary to obtain the first evaluation matrix.

[0069] Building a team of experts S ={ s 1, s 2,…, s m ,…, s M}, for the initial technology classification system T 0 for evaluation, where s m Indicates m Experts, m =1,2,…, M During the evaluation process, ensure M The evaluation experts are independent of each other and construct the following first evaluation matrix V :

[0070] ;

[0071] in, For the m Experts s m For i Technical Classification Vocabulary w i The weighted assessment of importance takes into account the subjective scores of experts and the importance of vocabulary. The calculation formula is as follows:

[0072] .

[0073] For the m Experts on i Technical Classification Vocabulary w i The evaluation of importance ranges from {unimportant, moderately important, important, very important, very important}, and the corresponding values ​​are 1, 3, 5, 7, 9; g’ ( w i ) is the i Technical Classification Vocabulary w iThe normalized importance of ; or is the weight coefficient, which is the weight of controlling the importance of the word. Considering that the highest expert score is 9, It is set to 5, which means that the influence of vocabulary importance on the final result is slightly weaker than that of expert scoring.

[0074] Step 5: Reduce the expert set according to the first evaluation matrix to obtain the second evaluation matrix, and calculate the third evaluation matrix through normalization.

[0075] Experts focus M Assessment experts S ={ s 1, s 2,…, s m ,…, s M} The degree of familiarity with the current assessment content varies. If the opinion of an expert differs greatly from that of the majority of experts, it is considered that the expert is not familiar with the current assessment content and needs to be eliminated.

[0076] No. m Deviation of the evaluation results of experts d m The calculation formula is:

[0077]

[0078] Where:

[0079] d m Indicates m Experts' familiarity with current assessment content, d m The larger the m The less familiar the experts are with the current assessment content, the lower the threshold of expert concentration reduction is. d 0∈(0,1]; if d j > d 0, the expert is deemed not familiar with the current assessment content and is removed; otherwise, he is retained.

[0080] The reduced expert set is , 1 ≤ r ≤ M .

[0081] Eliminate the first evaluation matrix V The row vector corresponding to the eliminated experts in the equation is obtained as follows: :

[0082]

[0083] After normalization, the third evaluation matrix is ​​obtained for:

[0084]

[0085] in: , i =1,2,…, N , m =1,2,…, r , v max The second evaluation matrix The maximum value of the weighted evaluation values ​​of the importance of technical classification terms by experts in v min It is the minimum value among the weighted evaluation values ​​of the importance of technical classification vocabulary by experts.

[0086] Step 6: Classify the importance and consistency of vocabulary based on the third evaluation matrix calculation technology.

[0087] Based on the third assessment matrix Comprehensive calculation of each technical classification vocabulary w i Importance v i :

[0088]

[0089] Standard deviation s Metrics for each technology classification vocabulary w i The consistency of the corresponding multiple evaluation results s i :

[0090] .

[0091] Step 7: Reduce the technical classification vocabulary set according to the importance and consistency of the technical classification vocabulary to obtain the final patent technology classification system.

[0092] Setting importance thresholds for technical taxonomy terms v 0∈(0,1) and the consistency threshold of the evaluation results s 0∈(0,1), when the following conditions are met, the technical classification vocabulary is considered unimportant and can be directly removed from the technical classification vocabulary set:

[0093] s i < s 0,v i < v 0

[0094] When the following conditions are met, the technical classification vocabulary is considered important and should be retained in the technical classification vocabulary set:

[0095] s i < s 0, v i ≥ v 0

[0096] when s i ≥ s 0, it indicates that the effectiveness of the evaluation result is poor and the choice of technical classification vocabulary cannot be determined. Return to step 4 for processing. Q After ( Q The value of is generally 3-5 times), if there are still controversial indicators, they will be deleted, and then the final technical classification system vocabulary set and the parent-child relationship of vocabulary will be formed, that is, the final technical classification system will be formed. T :

[0097]

[0098] This method adopts data-driven and group decision-making theory, and through domain corpus data mining and expert comprehensive evaluation, forms a patent technology classification construction method for a specific field that combines quantitative and qualitative methods, and has been effectively practiced in the construction of a patent technology classification system in the field of air traffic management.

[0099] Example:

[0100] Assuming that in order to construct the patent technology classification system in the field of air traffic management, the determined expert set is {expert 1, expert 2, expert 3, expert 4, expert 5}, the vocabulary set of the preliminary technology classification system formed in step 3 is arranged in descending order of importance, and the six words in the second level are {air traffic control, airspace management, traffic management, airport operation, air traffic control big data, air traffic control bureau}, and the expert scores are given to this set. The evaluation results of the five experts on each word are shown in Table 1.

[0101] Table 1 Expert evaluation results

[0102]

[0103] According to the evaluation results, according to step 5, the reference vector is constructed as follows:

[0104]

[0105] The deviation of each expert's evaluation opinion is as follows:

[0106]

[0107] Set the expert intensive reduction threshold , it can be seen that the deviation of expert 5 is greater than the threshold. It can be considered that the expert is not familiar with the current evaluation content and can be reduced from the expert set.

[0108] The evaluation results after expert reduction are shown in Table 2. Based on step 6 and comprehensive expert opinions, the importance of each word is as follows:

[0109]

[0110] At the same time, the evaluation consistency of each word is obtained:

[0111]

[0112] Table 2 Evaluation results after expert reduction

[0113]

[0114] Assume that the importance threshold of a word , evaluation result consistency threshold According to the judgment criteria of step 7, the four terms "air traffic control", "airspace management", "flow management" and "airport operation" can be accepted, and "air traffic control bureau" and "air traffic control big data" can be eliminated.

[0115] Furthermore, the 24 words in the third level were evaluated by experts, and finally the final index classification system was obtained, such as Figure 4 shown.

[0116] In a specific implementation, the present application provides a computer storage medium and a corresponding data processing unit, wherein the computer storage medium can store a computer program, and when the computer program is executed by the data processing unit, the invention content of the method for constructing a patent technology classification system in the field of air traffic management provided by the present invention and some or all of the steps in each embodiment can be run. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0117] Those skilled in the art can clearly understand that the technical solutions in the embodiments of the present invention can be implemented by means of computer programs and their corresponding general hardware platforms. Based on this understanding, the technical solutions in the embodiments of the present invention are essentially or the part that contributes to the prior art can be embodied in the form of a computer program, i.e., a software product, which can be stored in a storage medium and includes several instructions for enabling a device including a data processing unit (which can be a personal computer, a server, a single-chip microcomputer, a MUU or a network device, etc.) to execute the methods described in various embodiments of the present invention or certain parts of the embodiments.

[0118] The present invention provides a method for constructing a patent technology classification system in the field of air traffic management. There are many methods and ways to implement the technical solution. The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the protection scope of the present invention. All components not specified in this embodiment can be implemented by existing technologies.

Claims

1. A method for constructing a patent technology classification system in the field of air traffic management, characterized in that: The following steps are involved: Step 1: mining key technical classification vocabulary in the field of air traffic management to obtain a technical classification vocabulary set; Step 2, calculating the measurement value of the technical classification vocabulary in the technical classification vocabulary set; Step 3: Construct an initial technical classification system for patents in the field of air traffic management based on the measurement values ​​of technical classification vocabulary; Step 4, constructing an expert set, wherein the experts in the expert set evaluate the technical classification vocabulary to obtain a first evaluation matrix; Step 5, reducing the expert set according to the first evaluation matrix to calculate a third evaluation matrix; reducing the expert set includes judging whether the experts in the expert set are familiar with the technical classification vocabulary according to the first evaluation matrix, and removing experts who are not familiar with the current evaluation content from the expert set; Step 6, calculating the importance and consistency of the technical classification vocabulary according to the third evaluation matrix; Step 7, reducing the technical classification vocabulary set according to the importance and consistency of the technical classification vocabulary to obtain the final patent technology classification system; the reducing the technical classification vocabulary set includes judging whether the technical classification vocabulary is important according to the importance and consistency of the technical classification vocabulary, and removing unimportant technical classification vocabulary from the technical classification vocabulary set.

2. The method for constructing a patent technology classification system in the field of air traffic management according to claim 1 is characterized in that: Step 1 includes: Building a corpus in the field of air traffic management D ={ d 1, d 2,…, d k ,…, d K }, K For the corpus D Total number of documents in d k Representation Corpus D Middle k documents, 1 ≤ k ≤ K ; From the corpus D Extract keywords from the data and process them to form a technical classification vocabulary set in the field of air traffic management W ={ w 1, w 2,…, w i ,…, w N }, N is the total number of technical classification words, w i Vocabulary of technical classification W Middle i words, 1 ≤ i ≤ N .

3. The method for constructing a patent technology classification system in the field of air traffic management according to claim 2 is characterized in that: The measurement value of the technical classification vocabulary in step 2 includes the co-occurrence frequency of two words in the technical classification vocabulary set and the importance of the words; the co-occurrence frequency of the two words is calculated Co ( w i , w j )The formula is: , in, w i , w j Vocabulary of technical classification W The vocabulary in , 1 ≤ i , j ≤ N ; When the vocabulary w i and w j Also appears in the document d k middle, P k ( w i , w j )=1, otherwise P k ( w i , w j )=0; Computing vocabulary w i Importance g ( w i )The formula is: .

4. The method for constructing a patent technology classification system in the field of air traffic management according to claim 3 is characterized in that: Step 3 includes building the parent-child relationship of vocabulary to form an initial technical classification system: W Any word in w c , for which the parent node vocabulary is selected w p , if and only if ; In each child node w c and the parent node w p Add an edge between e cp , forming the initial technology classification system T 0={ W , E },in, .

5. The method for constructing a patent technology classification system in the field of air traffic management according to claim 4 is characterized in that: Step 4 involves building a team of experts S ={ s 1, s 2,…, s m ,…, s M }, for the initial technology classification system T 0 for evaluation, where s m represents the mth expert, m =1,2,…, M During the evaluation process, ensure M The evaluation experts are independent of each other and construct the following first evaluation matrix V : ， in, For the m Experts s m For i Technical Classification Vocabulary w i Weighted assessment of importance.

6. The method for constructing a patent technology classification system in the field of air traffic management according to claim 5 is characterized in that: Step 4 m Experts s m For i Technical Classification Vocabulary w i Weighted Assessment of Importance Taking into account the expert ratings and w i Importance g ( w i ), the calculation formula is: , For the m Experts on i Technical Classification Vocabulary w i The evaluation of importance ranges from {unimportant, moderately important, important, very important, very important}, and the corresponding values ​​are 1, 3, 5, 7, 9; g’ ( w i ) is the i Technical Classification Vocabulary w i The normalized importance of ; η is the weight coefficient, which is the weight for controlling the importance of vocabulary.

7. The method for constructing a patent technology classification system in the field of air traffic management according to claim 6 is characterized in that: Step 5 includes: judging whether the deviation of the comprehensive evaluation results of all experts is greater than the set threshold, and the deviation of the evaluation result of the mth expert δ m The calculation formula is: , In the formula, , δ m Indicates m The expert's familiarity with the current assessment content, the set expert concentration reduction threshold δ 0∈(0,1]; if δ j > δ 0, the expert is deemed not familiar with the current assessment content and is removed, otherwise retained; Thus, the reduced expert set is obtained as , 1 ≤ r ≤ M .

8. The method for constructing a patent technology classification system in the field of air traffic management according to claim 7 is characterized in that: Step 5 also includes: removing the row vectors corresponding to the removed experts in the first evaluation matrix V, and obtaining the second evaluation matrix: : ， After normalization, the third evaluation matrix is ​​obtained for: ， in: , i =1,2,…, N , m =1,2,…, r , v max The second evaluation matrix The maximum value of the weighted evaluation values ​​of the importance of technical classification terms by experts in v min It is the minimum value among the weighted evaluation values ​​of the importance of technical classification vocabulary by experts.

9. The method for constructing a patent technology classification system in the field of air traffic management according to claim 8 is characterized in that: Step 6 includes: Based on the third evaluation matrix Comprehensive calculation of each technical classification vocabulary w i Importance v i : ， The standard deviation is used to measure each technical classification vocabulary w i The consistency of the corresponding multiple evaluation results σ i : 。 10. The method for constructing a patent technology classification system in the field of air traffic management according to claim 9, characterized in that: Step 7 includes: Classifying vocabulary by technology w i Importance v i and consistency σ i Air Traffic Management Technology Classification Set W ={ w 1, w 2,…, w N } to reduce and set the importance threshold of technical classification vocabulary v 0∈(0,1) and the consistency threshold of the evaluation results σ 0∈(0,1), when σ i < σ 0, v i < v When it is 0, the technical classification vocabulary is considered unimportant and is directly removed from the technical classification vocabulary set; When satisfied σ i < σ 0, v i ≥ v When the value is 0, the technical classification vocabulary is considered important and is retained in the technical classification vocabulary set: when σ i ≥ σ 0, it indicates that the effectiveness of the evaluation result is poor and the choice of technical classification vocabulary cannot be determined. Return to step 4 for processing; if repeated execution Q If there are still controversial technical classification terms after this, they will be deleted to form the final technical classification vocabulary set. Parent-child relationship with vocabulary : ， Where 1 ≤ p ≤ N , Q The value of is 3~5; Final patent technology classification system .

Images