Navigation announcement key statement extraction method and device, equipment and storage medium

Abstract

The invention discloses a navigation announcement key statement extraction method, device and equipment and a storage medium, and relates to the technical field of aviation information, and the method comprises the steps: firstly obtaining a target navigation announcement; thirdly, performing sentence segmentation processing on the target navigation announcement, and constructing a navigation announcement statement set; then, based on initial weights of statements and similarities between the statements and statements in the initial candidate set, selecting target statements from the navigation announcement statement set to construct a final candidate set; and finally, based on a preset number of key statements, selecting a plurality of statements with the highest initial weight sequence from the final candidate set as key statements of the target navigation announcement. According to the embodiment of the invention, the key statements can be rapidly and accurately extracted from the navigation announcements, so that a worker can rapidly screen out useful navigation announcements, and omission and misjudgment are avoided.

Description

Technical Field

[0001] The present invention relates to the field of aviation information technology, and in particular to a method, device, equipment and storage medium for extracting key sentences from a NOTAM. Background Art

[0002] NOTAM (Notice To Airmen) is a notice sent by the aviation department to pilots or flight operation related units to deal with aviation emergencies. In addition to conveying relevant information, NOTAM is also the basis for arranging flight routes and flight times. According to statistics, the total number of NOTAMs sent worldwide each year is over one million, and the number is still increasing.

[0003] When processing navigation notices, staff often need to face multiple navigation notices to be processed at the same time, but navigation notices often contain a lot of information that is not directly related to the current task, which increases the difficulty for staff to screen information and reduces work efficiency; moreover, the long text content of navigation notices not only requires staff to spend a lot of reading time, but also easily causes reading fatigue and distraction of staff, which in turn leads to omission of key information and misjudgment, affecting navigation safety. Summary of the invention

[0004] The purpose of the embodiments of the present invention is to provide a method, device, equipment and storage medium for extracting key sentences from navigation notices, which can quickly and accurately extract key sentences from navigation notices, so that staff can quickly screen out useful navigation notices and avoid omissions and misjudgments.

[0005] To achieve the above object, an embodiment of the present invention provides a method for extracting key sentences from a navigation notice, comprising:

[0006] Obtain target NOTAMs;

[0007] Sentence processing is performed on the target NOTAM, and a NOTAM sentence set is constructed;

[0008] Selecting a target sentence from the NOTAM sentence set based on the initial weight of the sentence and its similarity to the sentences in the initial candidate set to construct a final candidate set;

[0009] Based on the preset number of key sentences, several sentences with the highest initial weight ranking are selected from the final candidate set as the key sentences of the target navigation notice.

[0010] As an improvement of the above scheme, the target sentence is selected from the navigation notice sentence set based on the initial weight of the sentence and its similarity with the sentences in the initial candidate set to construct a final candidate set, including:

[0011] Select the sentence with the largest initial weight from the NOTAM sentence set and add it to the initial candidate set;

[0012] Determining whether the number of sentences in the initial candidate set reaches a preset threshold;

[0013] If the number of sentences in the initial candidate set does not reach the preset threshold, the weight of the NOTAM of each sentence outside the initial candidate set is calculated, and the sentence with the largest NOTAM weight is selected to be added to the initial candidate set, and then the step of determining the number of sentences in the initial candidate set is returned; wherein the NOTAM weight is calculated based on the initial weight of each sentence and its similarity with the sentences in the initial candidate set;

[0014] If the number of sentences in the initial candidate set reaches the preset threshold, the current initial candidate set is used as the final candidate set.

[0015] As an improvement to the above solution, the weight of the NOTAM is calculated by the following formula:

[0016] W 1 (S i )=αW(S i )+sum[(α-1)σ(S i ,D)]

[0017] Among them, W 1 (S i ) represents the weight of the NOTAM of the i-th sentence; S i represents the i-th statement; W(S i ) represents the initial weight of the i-th sentence; α represents the adjustment coefficient; D represents the initial candidate set; σ(S i ,D) represents the statement S i The similarity with the sentences in the initial candidate set; sum represents the summation process.

[0018] As an improvement of the above solution, the initial weight is composed of sentence position weight, topic similarity weight, subject word weight and keyword weight.

[0019] As an improvement of the above solution, the initial weight is calculated in the following way:

[0020] Normalizing the sentence position weight, the topic similarity weight, the subject word weight, and the keyword weight respectively to obtain a first sentence position weight, a first topic similarity weight, a first subject word weight, and a first keyword weight;

[0021] The first sentence position weight, the first subject word weight and the first keyword weight are summed to obtain an intermediate weight;

[0022] Normalizing the intermediate weight to obtain a first intermediate weight;

[0023] The first intermediate weight and the first topic similarity weight are summed to obtain the initial weight.

[0024] As an improvement to the above solution, the sentence position weight is calculated by the following formula:

[0025]

[0026] Among them, W L (S i ) represents the sentence position weight of the i-th sentence; S i represents the i-th sentence; y represents the total number of sentences in the first paragraph of the target NOTAM; u represents the total number of sentences in the last paragraph of the target NOTAM; n represents the total number of sentences in the target NOTAM; z 1 and z 2 Represents the control parameters.

[0027] As an improvement to the above solution, the subject word weight is calculated by the following formula:

[0028]

[0029] Among them, W C (S i ) represents the topic word weight of the i-th sentence; S i Represents the i-th statement.

[0030] To achieve the above purpose, an embodiment of the present invention further provides a device for extracting key sentences from navigation notices, comprising:

[0031] A navigation notice acquisition module is used to obtain target navigation notices;

[0032] A sentence segmentation module, used for performing sentence segmentation processing on the target NOTAM and constructing a NOTAM sentence set;

[0033] A candidate set construction module, used for selecting a target sentence from the NOTAM sentence set based on the initial weight of the sentence and its similarity with the sentences in the initial candidate set, so as to construct a final candidate set;

[0034] The key sentence extraction module is used to select a number of sentences with the highest initial weight ranking from the final candidate set based on a preset number of key sentences as the key sentences of the target navigation notice.

[0035] To achieve the above-mentioned purpose, an embodiment of the present invention further provides a navigation notice key sentence extraction device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the navigation notice key sentence extraction method as described in any of the above-mentioned embodiments is implemented.

[0036] To achieve the above-mentioned purpose, an embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the method for extracting key sentences from the navigation notice as described in any of the above-mentioned embodiments.

[0037] Compared with the prior art, the method, device, equipment and storage medium for extracting key sentences from a navigation notice provided in the embodiment of the present invention obtain a target navigation notice; perform sentence processing on the target navigation notice and construct a navigation notice sentence set; based on the initial weight of the sentence and its similarity with the sentences in the initial candidate set, select the target sentence from the navigation notice sentence set to construct a final candidate set; based on the preset number of key sentences, select several sentences with the highest initial weight ranking from the final candidate set as the key sentences of the target navigation notice. The embodiment of the present invention can quickly and accurately extract key sentences from the navigation notice, so that the staff can quickly screen out useful navigation notices and avoid omissions and misjudgments. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a flow chart of a method for extracting key sentences of a navigation notice provided by one embodiment of the present invention;

[0039] Figure 2 is a flow chart of a method for extracting key sentences of a navigation notice provided by another embodiment of the present invention;

[0040] Figure 3 It is a structural schematic diagram of a key sentence extraction device for navigation notices provided by one embodiment of the present invention;

[0041] Figure 4 The present invention is a schematic diagram of the structure of a key sentence extraction device for navigation notices provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0042] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] In the description of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the feature. In the description of this application, unless otherwise specified, "plurality" means two or more.

[0044] See also Figure 1 , is a flow chart of a method for extracting key sentences of a navigation notice provided by an embodiment of the present invention, comprising steps S1 to S4:

[0045] S1. Obtain target navigation notice;

[0046] S2, processing the target NOTAM sentence by sentence, and constructing a NOTAM sentence set;

[0047] S3, based on the initial weight of the sentence and its similarity with the sentences in the initial candidate set, selecting the target sentence from the navigation notice sentence set to construct a final candidate set;

[0048] S4. Based on the preset number of key sentences, a number of sentences with the highest initial weight ranking are selected from the final candidate set as the key sentences of the target navigation notice.

[0049] Specifically, in step S2, the content of item E of the target navigation notice is subjected to syntactic analysis, sentence segmentation is implemented, and then each sentence is subjected to word segmentation, and stop words are removed to construct a navigation notice sentence set.

[0050] It is worth noting that, in order to evaluate the importance of different sentences more comprehensively and meticulously, the embodiment of the present invention calculates the initial weight of each sentence according to the influencing factors of the sentence importance, and sorts the initial weight of each sentence from large to small in step S4, and then selects several sentences with the highest ranking as the key sentences of the target navigation notice. For example, when the preset number of key sentences is 4, the top four sentences are selected as the key sentences. Exemplarily, after the key sentences are selected, the key sentences can also be combined to form a navigation notice summary.

[0051] Furthermore, the inventors of the present application have discovered through research that if key sentences are selected based only on the initial weight ranking, it may result in the selected key sentences having a high similarity and causing information redundancy. In order to avoid the above problems, in step S3, based on the initial weight of the sentence and its similarity with the sentences in the initial candidate set, the target sentence is selected and the final candidate set is constructed. That is, when constructing the candidate set, the embodiment of the present invention not only considers the initial weight of the sentence, but also considers the similarity between each sentence and the sentences in the initial candidate set, and performs a similarity penalty on the sentence, thereby reducing the similarity of each sentence in the final candidate set. Furthermore, in step S4, key sentences are selected within the final candidate set based on the initial weight ranking, so that the generalization and redundancy of each key sentence can be balanced.

[0052] As an optional implementation manner, the initial weight is composed of a sentence position weight, a topic similarity weight, a subject word weight and a keyword weight.

[0053] Compared with the prior art, the embodiment of the present invention can make a more comprehensive and detailed assessment of the importance of each sentence in the navigation notice by considering multiple types of weights such as sentence position weight, subject similarity weight, subject word weight and keyword weight, thereby improving the accuracy and representativeness of the extracted key sentences.

[0054] It is worth noting that the position of a sentence is one of the key factors in determining the importance of a sentence, especially the first and last paragraphs of a navigation notice, which often contain the summary content of the navigation notice. Therefore, the embodiment of the present invention calculates the sentence position weight according to the position of the sentence in the navigation notice. Specifically, in the first paragraph, the sentence position weight is negatively correlated with the "distance of the sentence from the first sentence", that is, as the sentence gradually moves away from the first sentence, the sentence position weight gradually decreases; while in the last paragraph, the sentence position weight is negatively correlated with the "distance of the sentence from the last sentence", that is, as the sentence gradually approaches the last sentence, the sentence position weight gradually increases.

[0055] As an optional implementation manner, the sentence position weight is calculated by the following formula:

[0056]

[0057] Among them, W L (S i ) represents the sentence position weight of the i-th sentence; S i represents the i-th sentence; y represents the total number of sentences in the first paragraph of the target NOTAM; u represents the total number of sentences in the last paragraph of the target NOTAM; n represents the total number of sentences in the target NOTAM; z 1 and z 2It represents a control parameter, which is used to adjust the sentence position weight of each sentence in the first and last paragraphs of the NOTAM. Its value range is usually limited to between 0 and 1, and the specific setting value needs to be adjusted and optimized according to the type of the target NOTAM.

[0058] Furthermore, the NOTAM code (Q-Code) is a code used in the NOTAM to identify the subject and status of the NOTAM content. See Table 1 for common NOTAM codes and their meanings provided by an embodiment of the present invention. In the NOTAM, the subject is a high-level summary and refinement of the NOTAM content. Therefore, the embodiment of the present invention uses subject similarity as an influencing factor for evaluating the importance of sentences, and assigns a higher subject similarity weight to sentences that are highly similar to the subject in content.

[0059] Specifically, the vectorization technology is used to convert the sentence and the topic into mathematical vector representations, and the cosine similarity between the vectors is used to measure the similarity between the sentence and the topic. For example, open source methods such as ERNIE (Wenxin Da Model), word2vec and BOW (Bag of Words) can be used to convert the sentence and the topic into vector representations. Suppose the sentence S i The vector representation of (a 1 ,...a m ), the vector of topic T is represented as (b 1 ,...b m ), then the topic similarity weight is calculated by the following formula:

[0060]

[0061] Among them, W T (S i ) represents the topic similarity weight of the i-th sentence; S i represents the i-th statement; a j Indicates S i The jth element in the vector representation of b j The jth element in the vector representation of the subject of the target NOTAM; m represents the dimension of the vector space.

[0062] Table 1 Common NOTAM codes and their meanings

[0063]

[0064] Furthermore, the number of subject words contained in a sentence is also a factor in determining the importance of the sentence, so the embodiment of the present invention also calculates the subject word weight of the sentence. Furthermore, since navigation notices often use abbreviations to describe the content, for example, the navigation code MRLC means RUNWAY CLOSED, but RUNWAY can be abbreviated to RWY, R / W, RNWY, RW and RY, etc., and CLOSED can be abbreviated to CL, CLDCLSD, etc., the embodiment of the present invention also counts the above subject abbreviations as subject words when calculating the subject word weight.

[0065] Furthermore, before calculating the subject word weight, it is also necessary to set a subject word set for each Q-Code subject, and the subject word set includes corresponding subject abbreviations.

[0066] As an optional implementation, the subject word weight is calculated by the following formula:

[0067]

[0068] Among them, W C (S i ) represents the topic word weight of the i-th sentence; S i Represents the i-th statement.

[0069] Furthermore, sentences containing text keywords usually have more text valid information than other sentences. Therefore, the embodiment of the present invention adopts the TF-IDF (Term Frequency-Inverse Document Frequency) algorithm to extract keywords in the target NOTAM and perform keyword audit. Specifically, the TF-IDF algorithm is used to extract keywords in the target NOTAM and construct a keyword set, wherein the size of the keyword set (the number of keywords) is positively correlated with the text length of the target NOTAM, and is adjusted and optimized according to the type of the target NOTAM.

[0070] Exemplarily, the keyword weight is calculated by the following formula:

[0071]

[0072] Among them, W k (S i ) represents the keyword weight of the i-th sentence; S i Represents the i-th statement.

[0073] As an optional implementation manner, the initial weight is calculated in the following manner:

[0074] T1, respectively normalizing the sentence position weight, the topic similarity weight, the subject word weight and the keyword weight to obtain a first sentence position weight, a first topic similarity weight, a first subject word weight and a first keyword weight;

[0075] T2, summing the first sentence position weight, the first subject word weight, and the first keyword weight to obtain an intermediate weight;

[0076] T3, normalizing the intermediate weight to obtain a first intermediate weight;

[0077] T4. Summing the first intermediate weight and the first topic similarity weight to obtain the initial weight.

[0078] It is worth noting that in order to eliminate the unfairness caused by dimensional differences between different types of weights, after calculating the sentence position weight, topic similarity weight, subject word weight and keyword weight, it is necessary to normalize each type of weight to ensure that the values ​​of each type of weight can be compared and analyzed under a unified measurement scale.

[0079] Exemplarily, in step T1, the Min-Max normalization method is used for normalization. This method can transform the original data into the interval [0,1] through linear transformation. The conversion formula is as follows:

[0080]

[0081] Among them, W NX (S i ) represents the normalized weight value; W X (S i ) represents the original weight value; W max (S i ) is the maximum value in the weight set; W min (S i ) represents the minimum value in the weight set.

[0082] Taking the sentence position weight as an example, when normalizing the sentence position weight, W X (S i ) is the sentence position weight W L (S i ), W NX (S i ) is the first sentence position weight W NL (S i ), W max (S i ) and W min (S i) are the maximum and minimum values ​​in the weight set, respectively, where the weight set is a set of sentence position weights of each sentence.

[0083] It is worth noting that, since the topic similarity weight is particularly important in the key sentence recognition process, the embodiment of the present invention regards it and the sum of the other three weights (intermediate weight) as equal weight. Specifically, in step T2, the first sentence position weight, the first subject word weight and the first keyword weight are summed to obtain the intermediate weight. The calculation process is shown in formula (6):

[0084] W MID (S i )=W NL (S i )+W NC (S i )+W NK (S i ) (6)

[0085] Among them, W MID (S i ) represents the statement S i The intermediate weight of NL (S i ) represents the statement S i The first sentence position weight of NC (S i ) represents the statement S i The first keyword weight of W NK (S i ) represents the statement S i The first keyword weight.

[0086] Further, in step T3, the intermediate weight is normalized to obtain the first intermediate weight W NM (S i ).

[0087] In step T4, the first intermediate weight and the first topic similarity weight are summed to obtain the initial weight. The calculation process is shown in formula (7):

[0088] W(S i )=W NT (S i )+W NM (S i ) (7)

[0089] Among them, W(S i ) represents the statement S i The initial weight of NT (S i ) represents the statement S iThe first topic similarity weight of W NM (S i ) represents the statement S i The first intermediate weight of .

[0090] Compared with the prior art, the embodiment of the present invention obtains an intermediate weight by summing the normalized sentence position weight, subject word weight and keyword weight, and then obtains an initial weight by summing the normalized intermediate weight and the topic similarity weight. This can balance the significant influence of topic similarity and the combined influence of other factors, so that the calculated initial weight can more accurately evaluate the importance of different sentences.

[0091] As an optional implementation, the step of selecting a target sentence from the NOTAM sentence set based on the initial weight of the sentence and its similarity to the sentences in the initial candidate set to construct a final candidate set (step S3) includes:

[0092] S31, selecting a sentence with the largest initial weight from the NOTAM sentence set and adding it to the initial candidate set;

[0093] S32, determining whether the number of sentences in the initial candidate set reaches a preset threshold;

[0094] S33, if the number of sentences in the initial candidate set does not reach the preset threshold, then calculating the navigation notice weight of each sentence outside the initial candidate set, and selecting the sentence with the largest navigation notice weight to add to the initial candidate set, and then returning to the step of determining the number of sentences in the initial candidate set; wherein the navigation notice weight is calculated based on the initial weight of each sentence and its similarity with the sentences in the initial candidate set;

[0095] S34: If the number of sentences in the initial candidate set reaches the preset threshold, the current initial candidate set is used as the final candidate set.

[0096] From the above content, it can be known that in order to avoid information redundancy in the selected key sentences, the embodiment of the present invention pre-constructs a candidate set and selects key sentences within the candidate set based on the initial weights, wherein the final candidate set is the candidate set ultimately used to extract the key sentences, and the initial candidate set is the intermediate candidate set (stage candidate set) generated in the process of constructing the final candidate set.

[0097] It is understandable that, at the beginning, the initial candidate set is an empty set, so in step S31, the sentence with the largest initial weight is directly selected from the navigation notice sentence set to construct the initial candidate set without performing a similarity penalty.

[0098] In step S32, it is determined whether the number of sentences in the initial candidate set reaches a preset threshold. If it does not reach the preset threshold, step S33 is executed to calculate the navigation notice weights of each sentence outside the initial candidate set and select the sentence with the largest navigation notice weight to add to the initial candidate set, and then return to step S32 for iteration; if it reaches the preset threshold, the iteration is stopped and the current initial candidate set is used as the final candidate set. The preset threshold is adjusted and optimized according to the navigation notice type and its characteristics to reflect the positive correlation between the total number of sentences in the final candidate set and the total number of navigation notice sentences.

[0099] As an optional implementation, the NOTAM weight is calculated by the following formula:

[0100] W 1 (S i )=αW(S i )+sum[(α-1)σ(S i ,D)] (8)

[0101] Among them, W 1 (S i ) represents the weight of the NOTAM of the i-th sentence; S i represents the i-th statement; W(S i ) represents the initial weight of the i-th sentence; α represents the adjustment coefficient; D represents the initial candidate set; σ(S i ,D) represents the statement S i The similarity with the sentences in the initial candidate set; sum represents the summation process.

[0102] In formula (8), the adjustment coefficient α is used to adjust the generalization and redundancy of the sentences in the initial candidate set, and (α-1) is a negative value. When there is potential duplication or redundant information between two sentences, a weight penalty is imposed on the sentence to be evaluated, that is, its weight is reduced. By reducing the weight, the accumulation of redundant information in the initial candidate set can be curbed, thereby minimizing redundancy while maintaining information density.

[0103] In order to more clearly understand the method of extracting key sentences from NOTAM, an example is given below to illustrate. Figure 2 , is a flowchart of a method for extracting key sentences of a navigation notice provided by another embodiment of the present invention, comprising Figure 2 It can be seen that the embodiment of the present invention first constructs a navigation notice statement set based on the target navigation notice, and then determines whether the candidate set is full. If the candidate set is not full, the initial weight and similarity penalty value of each sentence outside the candidate set are calculated, and the navigation notice weight of each sentence is calculated based on the initial weight and the similarity penalty value. After that, the sentence with the largest navigation notice weight is selected to add to the candidate set, and it is re-determined whether the candidate set is full, and the cycle is iterated until the candidate set is full; if the candidate set is full, the key sentence is output based on the candidate set.

[0104] Compared with the prior art, the method for extracting key sentences from a navigation notice provided in an embodiment of the present invention obtains a target navigation notice; performs sentence processing on the target navigation notice and constructs a navigation notice sentence set; selects a target sentence from the navigation notice sentence set based on the initial weight of the sentence and its similarity with the sentences in the initial candidate set to construct a final candidate set; based on a preset number of key sentences, selects several sentences with the highest initial weight ranking from the final candidate set as the key sentences of the target navigation notice. The embodiment of the present invention can quickly and accurately extract key sentences from the navigation notice, so that the staff can quickly screen out useful navigation notices and avoid omissions and misjudgments.

[0105] See also Figure 3 The embodiment of the present invention further provides a device 10 for extracting key sentences of navigation notices, comprising:

[0106] A navigation notice acquisition module 11 is used to acquire a target navigation notice;

[0107] A sentence segmentation module 12, used for segmenting the target NOTAM and constructing a NOTAM sentence set;

[0108] A candidate set construction module 13 is used to select a target sentence from the navigation notice sentence set based on the initial weight of the sentence and its similarity with the sentences in the initial candidate set to construct a final candidate set;

[0109] The key sentence extraction module 14 is used to select a number of sentences with the highest initial weight ranking from the final candidate set based on a preset number of key sentences as the key sentences of the target navigation notice.

[0110] The key sentence extraction device for navigation notices provided in the embodiment of the present invention can implement all the process steps of the key sentence extraction method for navigation notices described in the above embodiment. The functions of each module and unit in the device and the technical effects achieved are respectively the same as the functions and technical effects achieved by the key sentence extraction method for navigation notices described in the above embodiment, and the specific implementation method will not be repeated here.

[0111] See also Figure 4 The embodiment of the present invention further provides a navigation notice key sentence extraction device 20, comprising a processor 21, a memory 22, and a computer program stored in the memory 22 and configured to be executed by the processor 21. When the processor 21 executes the computer program, the steps in the above-mentioned navigation notice key sentence extraction method embodiment are implemented, for example Figure 1 or, the processor 21 implements the functions of each module in the above-mentioned device embodiments when executing the computer program.

[0112] The navigation notice key sentence extraction device may be a computing device such as a desktop computer, a notebook, a PDA, and a cloud server. The navigation notice key sentence extraction device may include, but is not limited to, a processor and a memory. Those skilled in the art may understand that the schematic diagram is only an example of the navigation notice key sentence extraction device and does not constitute a limitation on the navigation notice key sentence extraction device. It may include more or less components than shown in the figure, or combine certain components, or different components. For example, the navigation notice key sentence extraction device may also include input and output devices, network access devices, buses, etc.

[0113] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the navigation notice key sentence extraction device, and uses various interfaces and lines to connect various parts of the entire navigation notice key sentence extraction device.

[0114] The memory can be used to store the computer program and / or module, and the processor realizes various functions of the navigation notice key sentence extraction device by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function, etc.; the data storage area can store data created according to the use of the controller, etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Med ia Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0115] Wherein, if the module integrated in the navigation notice key sentence extraction device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned method embodiments can be implemented. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can include: any entity or device that can carry the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-On lyMemory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium.

[0116] Compared with the prior art, the key sentence extraction device, equipment and storage medium of the navigation notice provided in the embodiment of the present invention obtains the target navigation notice; performs sentence processing on the target navigation notice and constructs a navigation notice sentence set; selects the target sentence from the navigation notice sentence set based on the initial weight of the sentence and its similarity with the sentence in the initial candidate set to construct a final candidate set; based on the preset number of key sentences, selects several sentences with the highest initial weight ranking from the final candidate set as the key sentences of the target navigation notice. The embodiment of the present invention can realize the automatic extraction of the key sentences of the navigation notice, improve the processing efficiency and accuracy, and facilitate the staff to quickly screen out useful navigation notices to avoid omissions and misjudgments.

[0117] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for extracting key sentences from NOTAMs, characterized in that: include: Obtain target NOTAMs; Sentence processing is performed on the target NOTAM, and a NOTAM sentence set is constructed; Selecting a target sentence from the NOTAM sentence set based on the initial weight of the sentence and its similarity to the sentences in the initial candidate set to construct a final candidate set; Based on the preset number of key sentences, several sentences with the highest initial weight ranking are selected from the final candidate set as the key sentences of the target navigation notice.

2. The method for extracting key sentences from navigation notices according to claim 1, characterized in that: The method of selecting a target sentence from the NOTAM sentence set based on the initial weight of the sentence and its similarity to the sentence in the initial candidate set to construct a final candidate set includes: Select the sentence with the largest initial weight from the NOTAM sentence set and add it to the initial candidate set; Determining whether the number of sentences in the initial candidate set reaches a preset threshold; If the number of sentences in the initial candidate set does not reach the preset threshold, the weight of the NOTAM of each sentence outside the initial candidate set is calculated, and the sentence with the largest NOTAM weight is selected to be added to the initial candidate set, and then the step of determining the number of sentences in the initial candidate set is returned; wherein the NOTAM weight is calculated based on the initial weight of each sentence and its similarity with the sentences in the initial candidate set; If the number of sentences in the initial candidate set reaches the preset threshold, the current initial candidate set is used as the final candidate set.

3. The method for extracting key sentences of a NOTAM as claimed in claim 2, characterized in that: The NOTAM weight is calculated by the following formula: W1(S i )=αW(S i )+sum[(α-1)σ(S i ,D)] Among them, W1(S i ) represents the weight of the NOTAM of the i-th sentence; S i represents the i-th statement; W(S i ) represents the initial weight of the i-th sentence; α represents the adjustment coefficient; D represents the initial candidate set; σ(S i ,D) represents the statement S i The similarity with the sentences in the initial candidate set; sum represents the summation process.

4. The method for extracting key sentences from navigation notices according to claim 1, characterized in that: The initial weight is composed of sentence position weight, topic similarity weight, subject word weight and keyword weight.

5. The method for extracting key sentences of a navigation notice according to claim 4, characterized in that: The initial weights are calculated as follows: Normalizing the sentence position weight, the topic similarity weight, the subject word weight, and the keyword weight respectively to obtain a first sentence position weight, a first topic similarity weight, a first subject word weight, and a first keyword weight; The first sentence position weight, the first subject word weight and the first keyword weight are summed to obtain an intermediate weight; Normalizing the intermediate weight to obtain a first intermediate weight; The first intermediate weight and the first topic similarity weight are summed to obtain the initial weight.

6. The method for extracting key sentences from navigation notices according to claim 4, characterized in that: The sentence position weight is calculated by the following formula: Among them, W L (S i ) represents the sentence position weight of the i-th sentence; S i represents the i-th sentence; y represents the total number of sentences in the first paragraph of the target NOTAM; u represents the total number of sentences in the last paragraph of the target NOTAM; n represents the total number of sentences in the target NOTAM; z1 and z2 represent control parameters.

7. The method for extracting key sentences from navigation notices according to claim 4, characterized in that: The subject word weight is calculated by the following formula: Among them, W C (S i ) represents the topic word weight of the i-th sentence; S i Represents the i-th statement.

8. A device for extracting key sentences from navigation notices, characterized in that: include: A navigation notice acquisition module is used to obtain target navigation notices; A sentence segmentation module, used for performing sentence segmentation processing on the target NOTAM and constructing a NOTAM sentence set; A candidate set construction module, used for selecting a target sentence from the NOTAM sentence set based on the initial weight of the sentence and its similarity with the sentences in the initial candidate set, so as to construct a final candidate set; The key sentence extraction module is used to select a number of sentences with the highest initial weight ranking from the final candidate set based on a preset number of key sentences as the key sentences of the target navigation notice.

9. A device for extracting key sentences from navigation notices, characterized in that: The invention comprises a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the method for extracting key sentences of navigation notices according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the method for extracting key sentences of navigation notices according to any one of claims 1 to 7.

Images