A battlefield information prioritization method and apparatus

Abstract

The application provides a battlefield information priority sequencing method and device, and belongs to the technical field of information processing. The method comprises the following steps: screening and extracting attributes and information of historical battlefield data, and constructing a source case library by using a key success factor method; extracting a similar case set from the source case library according to the similarity between a target case situation analysis report and a source case situation analysis report by using a vector space model; screening an optional case most similar to the target case from the similar case set according to a feature attribute set of the source case; and taking a battlefield information priority sequencing result of the optional case as a current battlefield information priority sequencing result of the target case. The application proposes an information priority sequencing method based on C-KVC, and improves the rationality and accuracy of battlefield information priority sequencing.

Description

Technical Field

[0001] This invention relates to the field of information processing technology, and in particular to a method and apparatus for prioritizing battlefield information. Background Technology

[0002] In informationized battlefield conditions, command activities need to address the problem of extracting high-priority information from massive amounts of data, allocating more resources to corresponding activities to maximize benefits. Therefore, research on information prioritization is of great value and significance. Information prioritization is a comprehensive evaluation problem involving multiple domains, levels, and aspects of uncertainty. Current main methods include case study decision-making, analytic hierarchy process (AHP), neural network methods, grey analysis, and Bayesian network inference.

[0003] Due to the sheer volume of battlefield information and the complex, uncertain relationships between it, accurately describing and prioritizing information using precise mathematical models is extremely difficult. Instead, historical data and expert experience are primarily used to address current challenges. Furthermore, since information prioritization is ultimately determined by decision-makers, differences in their knowledge backgrounds and operational philosophies can lead to varying prioritizations even in the same operational scenarios.

[0004] Therefore, how to prioritize battlefield information more reasonably and accurately based on the decision-maker's style and the characteristics of combat activities is an urgent problem to be solved. Summary of the Invention

[0005] This invention provides a method and apparatus for prioritizing battlefield information, which addresses the shortcomings of existing technologies that make it difficult to reasonably and accurately prioritize battlefield information based on the decision-maker's style characteristics and the characteristics of combat activities.

[0006] In a first aspect, the present invention provides a method for prioritizing battlefield information, comprising: filtering and extracting attributes and information from historical battlefield data; constructing a source case library using the critical success factor method; the historical battlefield data including historical battle data and training activity data; extracting a set of similar cases from the source case library based on the similarity between the situation analysis report of the target case and the situation analysis report of the source case using a vector space model; selecting candidate cases most similar to the target case from the set of similar cases based on the feature attribute set of the source cases; the feature attribute set including decision-maker style characteristics and combat activity characteristics; and using the battlefield information priority ranking result of the candidate cases as the current battlefield information priority ranking result of the target case; the source cases in the source case library are described using a quadruple representation; the quadruple representation is as follows:

[0007] C i ={Ri ,F i ,L i ,T i |i∈N};

[0008] Among them, C i F represents the i-th source case in the source case set. i Let L represent the set of feature attributes corresponding to the i-th source case. i T represents the information priority ranking result corresponding to the i-th source case. i Let N represent the trust factor of the sorting result corresponding to the i-th source case, and N represent the number of source cases in the case library; and,

[0009] F i =(f ij |j=1,2,…,m);

[0010] Among them, f ij Let m represent the j-th feature attribute of the i-th source case, where m is the dimension of the feature attribute set; the feature attributes include decision-maker style features and operational activity features.

[0011] According to a battlefield information priority ranking method provided by the present invention, after taking the battlefield information priority ranking result of the candidate cases as the current battlefield information priority ranking result of the target case, the method further includes: modifying the source cases and generating new source cases based on the ranking effect and the decision-maker's evaluation feedback information on the ranking result, so as to update the source case library.

[0012] According to a battlefield information priority ranking method provided by the present invention, the method utilizes a vector space model to extract a set of similar cases from the source case library based on the similarity between the situation analysis reports of target cases and the situation analysis reports of source cases. This includes: obtaining target keyword vectors from the situation analysis reports of target cases and keyword vectors from the situation analysis reports of each source case; obtaining the similarity between each source case and the target case based on each keyword vector and the target keyword vector; and extracting a set of similar cases from the source case library based on the similarity between each source case and the target case and a preset similarity threshold.

[0013] According to a battlefield information priority ranking method provided by the present invention, the step of selecting the most similar candidate case to the target case from the set of similar cases based on the feature attribute set of the source case includes: determining multiple target feature attributes that can better characterize the features of the source case from the set of feature attributes; determining the combined weight of each target feature attribute using a combination of subjective and objective weighting methods; calculating the similarity of each target feature attribute between each source case and the target case using different methods according to the type of the target feature attribute; calculating the feature similarity between each source case and the target case based on the similarity of each target feature attribute between each source case and the target case, and the combined weight of each target feature attribute; and determining the most similar source case to the target case as the candidate case based on the trust factor of each source case and the feature similarity.

[0014] According to a battlefield information priority ranking method provided by the present invention, before extracting a set of similar cases from the source case library based on the similarity between each source case and the target case and a preset similarity threshold, the method further includes: determining the maximum similarity and the minimum similarity from the similarity between each source case and the target case; and determining the preset similarity threshold based on the maximum similarity and the minimum similarity.

[0015] According to a battlefield information priority ranking method provided by the present invention, the method of determining the combined weight of each target feature attribute by using a combination of subjective and objective weighting methods includes: determining the subjective weight vector and objective weight vector of the target feature attribute; and determining the target weight vector of the target feature attribute based on the standard deviation of the subjective weight vector, the standard deviation of the objective weight vector, the subjective weight vector, and the objective weight vector.

[0016] According to a battlefield information priority ranking method provided by the present invention, the step of modifying source cases and generating new source cases to update the source case library based on the ranking effect and the decision-maker's evaluation feedback information on the ranking results includes: writing new cases based on the ranking results and the ranking effect corresponding to the ranking results, and adding the new cases to the source case library; modifying the trust factor of the candidate cases based on the decision-maker's evaluation feedback information to update the original data of the candidate cases.

[0017] Secondly, the present invention also provides a battlefield information priority sorting device, comprising:

[0018] The source case library construction module is used to filter and extract attributes and information from historical battlefield data, and to construct the source case library using the critical success factor method; the historical battlefield data includes historical battle data and training activity data.

[0019] The similar case set extraction module is used to extract a similar case set from the source case library based on the similarity between the situation analysis report of the target case and the situation analysis report of the source case using a vector space model;

[0020] The alternative case extraction module is used to select the alternative cases most similar to the target case from the set of similar cases based on the feature attribute set of the source case; the feature attribute set includes decision-maker style characteristics and combat activity characteristics;

[0021] The sorting result determination module is used to take the battlefield information priority sorting result of the candidate cases as the current battlefield information priority sorting result of the target case;

[0022] The source cases in the source case library are described using a quadruple representation; the quadruple representation is as follows:

[0023] C i ={R i ,F i ,L i ,T i |i∈N};

[0024] Among them, C i F represents the i-th source case in the source case set. i Let L represent the set of feature attributes corresponding to the i-th source case. i T represents the information priority ranking result corresponding to the i-th source case. i Let N represent the trust factor of the sorting result corresponding to the i-th source case, and N represent the number of source cases in the case library; and,

[0025] F i =(f ij |j=1,2,…,m);

[0026] Among them, f ij Let m represent the j-th feature attribute of the i-th source case, where m is the dimension of the feature attribute set; the feature attributes include decision-maker style features and operational activity features.

[0027] Thirdly, the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of any of the battlefield information priority sorting methods described above.

[0028] Fourthly, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the battlefield information priority sorting method described above.

[0029] The battlefield information prioritization method and apparatus provided by this invention combine Case Decision Theory (CBDT), Key Success Factors (KSF), Vector Space Model (VSM), and Case Reasoning (CBR) to extract historical battle cases or training exercises with similar operational scenarios to the current operational activities faced by the decision-maker. Simultaneously, based on the decision-maker's style and the characteristics of the specific operational activities, it selects the historical battle cases or training exercises most similar to the current operational activities, and uses their information priority ranking results as the current operational activity's information priority ranking results. This invention proposes a C-KVC-based information priority ranking method, improving the rationality and accuracy of battlefield information priority ranking. Attached Figure Description

[0030] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0031] Figure 1 This is a schematic diagram of the research process of the case reasoning technology provided by the present invention;

[0032] Figure 2 This is a flowchart illustrating the battlefield information priority ranking method based on C-KVC provided by the present invention.

[0033] Figure 3 This is a schematic diagram comparing the battlefield information prioritization based on C-KVC and human brain thinking, provided by the present invention.

[0034] Figure 4 This is a schematic diagram of the keyword extraction results in the target case situation analysis report provided by the present invention;

[0035] Figure 5 This is a diagram illustrating the text similarity between the target case and the source case analysis report provided by this invention;

[0036] Figure 6 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0037] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0038] It should be noted that, in the description of the embodiments of the present invention, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0039] Before introducing the technical solution of the present invention, the theoretical basis of the method used in the present invention will be briefly explained here.

[0040] Case Decision Theory

[0041] Case-Based Decision Theory (CBDT), first proposed by Gilboa and Schmeidler in their paper "Case-Based Decision Theory," is based on the idea of ​​using past similar problems and their solutions as a basis for predicting solutions to new problems. The problem of prioritizing battlefield information aligns with CBDT. In specific combat operations, the question of which information to recommend to decision-makers is crucial. Drawing on the main ideas of CBDT, the current operational problem faced by decision-makers is considered the target case. The prioritization results of information proposed by decision-makers with similar styles in similar combat scenarios are extracted and recommended to decision-makers in the target case.

[0042] Key Success Factors Method

[0043] The Key Success Factors (KSF) methodology emerged in the 1960s, originating from research on identifying organizational goals and strategies. Its aim is to identify the key factors that play a decisive role in achieving those goals and manage them to improve performance. In the construction of a source case library within information prioritization methods, KSF is a relatively feasible approach, with advantages primarily reflected in the following aspects: First, it has a clear hierarchy. From the initial identification of mission objectives to the determination of specific indicators and the formation of an evaluation system, KSF is a progressively refined and specific process, effectively addressing the difficulty of defining the scope of battlefield information. Second, it is highly targeted. KSF is tailored to specific combat styles and adversaries, making it more targeted. Third, it has strong compatibility. KSF can be combined with various methods such as interviews, and can adjust the specific indicators of key success factors based on internal and external changes and shifts in the enemy's and our own situation.

[0044] Vector Space Model

[0045] The Vector Space Model (VSM), proposed by Gerard Salton and successfully applied in information retrieval, is a commonly used method for text similarity analysis. The basic idea of ​​VSM-based text similarity analysis is to convert text into vector form, mapping text similarity calculations to a vector space, and then analyzing text similarity based on the similarity of vectors in the vector space. This study uses VSM to perform text similarity analysis on situation analysis reports generated from operational planning, thereby mapping the similarity between target and source case operational scenarios. By selecting an appropriate threshold, source cases similar to the target case's operational scenario are extracted, constructing a set of similar cases.

[0046] Case-based reasoning techniques

[0047] Case-Based Reasoning (CBR) is similar in concept to Case-Based Reasoning Technique (CBDT). When faced with a new problem, the brain often recalls similar cases from the past, comparing the context and timing of the new and old problems. Solutions to older cases are then adjusted and modified, applying past experience and methods to solve similar cases to address the current problem. The research process primarily includes case retrieval, case reuse, case adjustment, and case learning. Figure 1 The diagram shown is a schematic diagram of the case reasoning technology research process provided by the present invention.

[0048] This invention, based on the aforementioned theoretical foundation, proposes a battlefield information prioritization method based on C-KVC. The method includes, but is not limited to, the following steps: filtering and extracting attributes and information from historical battlefield data; constructing a source case library using the critical success factor method; extracting a set of similar cases from the source case library based on the similarity between the situation analysis reports of the target case and the source case using a vector space model; selecting candidate cases most similar to the target case from the set of similar cases based on the characteristic attribute set of the source cases; and using the battlefield information priority ranking results of the candidate cases as the current battlefield information priority ranking result of the target case. Furthermore, the source cases can be modified and new source cases generated based on the ranking effect and the decision-maker's evaluation feedback to update the source case library.

[0049] The following is combined Figures 2-6 This invention describes the technical solution of the battlefield information priority sorting method provided in the embodiments of the present invention. Figure 2 This is a flowchart illustrating the battlefield information priority ranking method based on C-KVC provided by the present invention. The seven steps in the diagram are explained below.

[0050] Step ①: Drawing on the CBDT concept, construct an overall framework for prioritizing battlefield information based on C-KVC.

[0051] Step 2: Collect historical battle examples and training activities, filter and extract attributes and information, and use KSF to build a source case library to provide a data foundation for prioritizing battlefield information.

[0052] Step 3: Use the VSM method to analyze the similarity between the target case and the situation analysis report of the source case library, extract the similar case set, that is, screen the source cases that are similar to the combat scenarios of the target case, and provide source cases with similar combat scenarios for priority ranking of battlefield information.

[0053] Step 4: In the set of similar cases with similar combat scenarios to the target case, considering the decision-maker's style characteristics and specific combat activity characteristics, the ideas and methods of case similarity analysis in CBR are used to select and generate the candidate cases most similar to the target case. Specifically, this mainly involves reducing the case attribute space based on methods such as deviation scores, determining the case attribute weights based on methods such as binary semantic operators and information entropy, and using case similarity analysis methods to calculate the similarity between each source case and the target case in the similar case set.

[0054] Step 5: Determine the priority ranking result of battlefield information in the candidate cases as the priority ranking result of the current battlefield information.

[0055] At this point, the battlefield information priority sorting process has been completed. The following steps ⑥ and ⑦ mainly concern the updating and improvement of the source case library, in order to form an "incremental learning" source case update mechanism and gradually improve the accuracy and adaptability of the sorting.

[0056] Step 6: Based on the ranking results and the decision-makers' feedback on the ranking results, modify the source cases and generate new source cases to achieve self-updating and improvement of the source case library.

[0057] Step 7: The dashed line indicates that there is no specific practical step. This means that in the next C-KVC-based battlefield information priority sorting practice, the updated source case library will be used.

[0058] Figure 3 This is a schematic diagram comparing the battlefield information prioritization based on C-KVC and human brain thinking, as provided by the present invention. Figure 3 As shown, the battlefield information prioritization method based on C-KVC mainly imitates the human thinking pattern when solving problems, and demonstrates the overall application framework of the battlefield information prioritization method based on C-KVC.

[0059] It can be seen that human thinking is limited to personal historical experience and knowledge structure, while the battlefield information prioritization framework allows decision-makers to draw on their own experience and knowledge structure while also learning from the experience of other decision-makers with similar styles, thereby improving the comprehensiveness and accuracy of the prioritization.

[0060] The following example, using a specific battlefield information priority ranking method, further illustrates the implementation of the battlefield information priority ranking method proposed in this invention.

[0061] Based on the above embodiments, as an optional embodiment, the battlefield information priority sorting method provided by the present invention for constructing a source case library is implemented as follows.

[0062] The paradigmatic description of a case needs to be combined with battlefield activities and is the basis for prioritizing information. The paradigmatic representation of a case generally needs to include three elements: "problem," "solution," and "solution implementation effect." Therefore, it can be represented using a four-tuple format: <Situation Analysis Report (R), Feature Attributes (F), Ranking Result (L), Trust Factor (T)>. A single source case is represented as C. i ={R i ,F i ,L i ,T i |i∈N},C i Let N be the i-th source case in the source case set, and let N represent the number of source cases in the case library.

[0063] First-dimensional situation analysis report (R). i This represents the current battlefield situation text report corresponding to the i-th source case, which is compiled and summarized based on historical battle cases and training data, and is used to compare the similarity of battlefield environments.

[0064] Situation analysis reports are conclusions drawn from analyzing information gathered through various channels regarding enemy, friendly, and battlefield environments. The collection and organization of these reports is the starting point and crucial step in building a source case library. The quantity and quality of these reports determine the overall quality of the source case library. First, the scope of situation analysis report collection must be expanded. During the collection process, as many similar reports as possible should be collected based on the scenarios that may be encountered in the current combat operations, ensuring the universality of the source case library. Second, the quality of the situation analysis reports must be ensured. As a type of operational document, situation analysis reports generally have a standardized format, reflect the combat scenario at the time, and serve as an important basis for subsequent steps. Therefore, it is essential to ensure that the collected situation analysis reports have a consistent format and complete structure.

[0065] Secondary feature attribute (F). F i =(f ij |j=1,2,…,m),F i Let f represent the set of operational activity feature attributes corresponding to the i-th source case. ij Let m represent the j-th feature attribute of the i-th source case, and m be the dimension of the feature attribute set.

[0066] Characteristic attributes include decision-maker style characteristics and operational activity characteristics. They are primarily used in later case similarity calculations to generate candidate cases and extract their ranking results. Therefore, the accuracy of the characteristic attributes determines the quality of the ranking results. Regarding characteristic attributes, it is essential to comprehensively summarize and establish a complete set of characteristic attributes that may influence operational planning methods and directions. Characteristic attributes generally involve two aspects: determining the attribute space and determining the attribute values. Determining the attribute space presents significant uncertainty due to the diverse attributes of operational activities, with different activities generally corresponding to different characteristic attributes. This chapter adopts the KSF method for determination. For determining specific attribute values, such as the decision-maker's age and operational philosophy, historical case studies and various training materials often only contain the commander's name; therefore, further collection and organization are needed to confirm these values.

[0067] Third-order demand list (L). i This represents the information priority sorting result corresponding to the i-th source case.

[0068] Determining the priority ranking of battlefield information is a crucial and conclusive outcome in building the source case library. It is also the most complex and labor-intensive step, requiring the identification of characteristic attributes under various operational scenarios. When generating the ranking results, it is essential to clarify the individual characteristics of decision-makers, differentiate between different operational scenarios, and select as many experts as possible for questionnaires and interviews as possible to represent various command styles, ensuring the diversity of the case library's personalized requirements list.

[0069] The fourth element of trust (T). i This represents the credibility of the ranking result corresponding to the i-th source case, and also the effectiveness of the solution implementation for the requirement generation problem. The initial trust factor value can be defined based on the specific source of the source case, and subsequent updates and improvements are mainly determined based on the decision-makers' feedback on the ranking results.

[0070] Based on the above embodiments, as an optional embodiment, the battlefield information priority ranking method provided by the present invention, which utilizes a vector space model to extract a similar case set from the source case library based on the similarity between the situation analysis report of the target case and the situation analysis report of the source case, includes: obtaining the target keyword vector of the situation analysis report of the target case, and the keyword vector of the situation analysis report of each source case; obtaining the similarity between each source case and the target case based on each keyword vector and the target keyword vector; and extracting a similar case set from the source case library based on the similarity between each source case and the target case and a preset similarity threshold.

[0071] Specifically, it includes the following steps:

[0072] Step 1: Preprocessing Situation Analysis Reports (R). The 21 collected situation analysis reports for maritime operations (20 source cases and 1 target case) are segmented and stop words are removed to transform the text reports into a series of words or morphemes.

[0073] Step 2: Extract keywords for the situation analysis report. Keywords are extracted using TF-IDF (Term Frequency-Inverse Document Frequency). The core idea is that the importance of a word is not only directly proportional to the number of times it appears in the text, but also inversely proportional to its frequency in the corpus. After calculating the "term frequency (TF)" and "inverse document frequency (IDF)" of the morphemes, the TF-IDF value is calculated using the formula TFIDF = TF × IDF. A higher TF-IDF value indicates greater importance to the text; therefore, an appropriate threshold can be set to identify morphemes above the threshold as keywords representative of the report text. Figure 4 This is a schematic diagram of the keyword extraction results in the target case situation analysis report provided by the present invention.

[0074] Step 3: Calculate the similarity of combat scenarios. The cosine similarity algorithm is used to calculate the similarity of combat scenarios. Compared to distance metrics, cosine similarity places greater emphasis on the directional differences between two text vectors. For example, the n-dimensional keyword vector (i.e., the target keyword vector) of the situation analysis report for target case X is represented as Vec(X) = (x1, x2, ..., x...). n ), where x i This represents the TF-IDF value of the keyword at position i, where i∈n; the m-dimensional keyword vector in the analysis report of a source case Y is Vec(Y)=(y1,y2,…,y m ), where y j Let represent the TF-IDF value of the keyword at position j, where j∈m; since the cosine value cos(θ) ranges between [-1,1], after normalization, the formula for calculating the similarity between the target case X and the source case Y in the combat scenario is:

[0075]

[0076] The closer the cosine similarity value (cosineSIM) is to 1, the closer the angle between the two vectors is to 0 degrees, meaning the two vectors are more similar in direction and represent more similar situation analysis reports. Conversely, the closer the angle is to 1, the closer the angle is to 0 degrees and the less similar the situation analysis reports represent. The text vector similarity between the target case situation analysis report and the source case situation analysis report is calculated, resulting in 20 similarity results. Figure 5 This is a diagram illustrating the text similarity between the target case and the source case analysis report provided by this invention.

[0077] Step 4: Set a similarity threshold to extract a set of similar cases. Based on the calculated similarity of the combat scenarios, a simple majority rule can be used to set a similarity threshold τ to filter the source cases, extracting the source cases whose similarity values ​​are greater than the similarity threshold τ to form a set of similar cases for the combat scenarios.

[0078] Optionally, the step of determining the similarity threshold includes: determining the maximum similarity and minimum similarity from the similarity between each source case and the target case; and determining the preset similarity threshold based on the maximum similarity and the minimum similarity.

[0079] The specific formula for calculating the similarity threshold τ is as follows:

[0080]

[0081] Among them, Sim (+) This represents the maximum value in the similarity calculation results, i.e., the maximum similarity:

[0082] Sim (+) =max{sim(R0,R i )|i∈N}

[0083] Sim (-) This represents the minimum value in the similarity calculation results, i.e., the minimum similarity:

[0084] Sim (-) =min{sim(R0,R i )|i∈N}

[0085] sim(R0,R i The number ) indicates the similarity between the situation analysis reports of the target case and the source case.

[0086] Using a simple majority rule, based on the above formula and similarity calculation results, the threshold τ = 0.37 is calculated. Therefore, C is selected. 17 ,C1,C 14 C7,C 10 This is a set of similar cases.

[0087] Based on the above embodiments, as an optional embodiment, the battlefield information priority ranking method provided by the present invention, wherein the step of selecting the most similar candidate case to the target case from the set of similar cases according to the feature attribute set of the source case, includes: determining multiple target feature attributes from the feature attribute set that can better characterize the features of the source case; determining the combined weight of each target feature attribute using a combination of subjective and objective weighting methods; calculating the similarity of each target feature attribute between each source case and the target case using different methods according to the type of the target feature attribute; calculating the feature similarity between each source case and the target case based on the similarity of each target feature attribute between each source case and the target case, and the combined weight of each target feature attribute; and determining the most similar source case to the target case as the candidate case based on the trust factor of each source case and the feature similarity.

[0088] Specifically, from the extracted set of cases with similar operational backgrounds, the single case with the most similar operational activity characteristics is selected and recorded as the preferred case (candidate case), further determining the priority ranking of battlefield information. The specific steps are as follows:

[0089] Step 1: Reduce the feature attribute space. Based on the characteristics of feature attribute values, they can be divided into five types: clear symbol type, precise numerical type, interval variable type, evaluative language type, and text type. Differentiate and select the deviation score calculation formula for each type, and reduce the operational feature attribute space based on the calculation results. The larger the value, the more the feature attribute reflects the characteristics of the case and should be retained; conversely, it can be reduced. Example: Select the top 10 attributes with the largest attribute deviation scores as the target feature attributes of the case, as shown in Table 1.

[0090] Table 1. Target Feature Attribute Information of Target Cases and Similar Case Sets

[0091]

[0092] Step 2: Determine the combined weights of the target feature attributes. A combined subjective and objective weighting method is used to determine the weights of the remaining feature attributes after reduction. Assume the attribute space is reduced from m dimensions to m' dimensions after reduction, and the subjective weights are W' = (w′1, w′2, ..., w′...). m' The objective weights are W"=(w″1,w″2,…,w″). m' Using the standard deviation ratio combined weighting method, subjective and objective weights are merged into a combined weight. The formula for calculating the combined weight of the target feature attribute is:

[0093] W = (w1, w2, ..., w m' )=θ′·W'+θ″·W"

[0094] in σ′ is the standard deviation of the subjective weight vector, and σ″ is the standard deviation of the objective weight vector.

[0095] According to Table 1, we can select F1, F3, F4, F9, F 10 F 13 F 15 F 16 F 24 F 26 Ten attributes were selected as target feature attributes for case similarity analysis. First, five experts were interviewed. Assuming each expert had equal weight, the subjective weights of the ten attributes were calculated using the subjective weighting method: (0.1, 0.08, 0.12, 0.08, 0.1, 0.14, 0.08, 0.09, 0.12, 0.09). Then, according to five information types (clear symbolic, precise numerical, interval variable, evaluative language, and textual), information entropy theory was used to obtain the objective weights: (0.12, 0.09, 0.12, 0.09). The values ​​0.17, 0.19, 0.09, 0.22, 0.26, 0.13, 0.13, 0.25, 0.2) can be normalized to (0.07, 0.1, 0.11, 0.05, 0.13, 0.15, 0.07, 0.07, 0.14, 0.11); finally, the combined weights are (0.08, 0.09, 0.12, 0.07, 0.12, 0.13, 0.08, 0.08, 0.13, 0.1).

[0096] Step 3: Calculate the similarity of operational features. When calculating the similarity of operational features, different similarity calculation formulas need to be used for different types of attributes. Let S... j (C0,C i () indicates similar combat scenarios, case C i The similarity to the target case C0 regarding attribute j has been mentioned in the construction of the source case library. ij Case C i The value of attribute j in f is then... 0j This represents the value of attribute j in the target case C0.

[0097] (1) If the attribute is a clear symbol, the attribute values ​​can be directly compared. The formula for calculating attribute similarity is:

[0098]

[0099] (2) For attributes that are precise numerical values, similarity can be calculated using distance metrics. To represent attribute values and The formula for calculating the distance between them is:

[0100] in

[0101] Then, the negative exponential function is used to convert it into a similarity score. The formula for calculating attribute similarity is:

[0102]

[0103] (3) If the attribute is an interval variable, similarity can also be calculated using distance metrics, such as... The formula for calculating attribute similarity is:

[0104]

[0105] (4) If it is an evaluation of linguistic attributes, the fuzzy evaluation linguistic variable can be first converted into a triangular fuzzy number, and then the similarity can be calculated using the distance measure method. Let... For an ordered set of evaluation phrases U j The (r+1)th evaluation phrase, U j There are a total of T j A few evaluation phrases, then It can be represented as a triangular fuzzy number The conversion formula is:

[0106]

[0107] set up Represents the evaluation language attribute value and The degree of difference between them is converted into triangular fuzzy numbers. and The formula for calculating attribute similarity is:

[0108]

[0109] (5) If it is a text attribute, a keyword set can be used. and The similarity is represented by the ratio of the intersection to the union of the two sets of attributes. The formula for calculating attribute similarity is:

[0110]

[0111] According to the formula above, the source case C can be obtained. i Similarity with target case C0 on m' primary attributes {S1(C0,C i ),S2(C0,C i ...S m' (C0,C i )}, combined with the combined weights W=(w1,w2,…,w m' From this, we can obtain similar source cases C of the combat scenario. iSimilarity in operational characteristics to target case C0:

[0112]

[0113] S(C0,C i )∈[0,1],S(C0,C i The closer the value is to 1, the better it represents source case C. i The higher the similarity to the target case C0, the lower the similarity, and vice versa. Therefore, the similarity between the target case C0 and the source cases C1, C7, and C8 is calculated respectively. 10 C 14 C 17 Similarity: S(C0,C1)=0.8, S(C0,C7)=0.86, S(C0,C7)=0.86, 10 )=0.90,S(C0,C 14 )=0.85,S(C0,C 17 = 0.82.

[0114] Step 4: Generate information priority ranking results. This involves obtaining the set of similar cases for all combat scenarios {C′1, C′2, ..., C′...}. n′ The similarity values ​​S1, S2, ..., S with the target case C0 n Based on this, a trust factor T is added, and the similarity values ​​S′1, S′2, ..., S′ after adding the trust factor are calculated. n The calculation formula is as follows:

[0115] S′ n =T·S n

[0116] Arrange them in descending order, and denote them as S′. (1) ≥S′ (2) ≥…≥S′ (j) ≥…≥S′ (n') At this point, the case with the highest similarity value is S′. (1) The information priority ranking result in the corresponding source case is the information priority ranking result of the target case.

[0117] Since the similar source cases in this practice were all constructed based on the critical success factor method, C1, C7, and C 10 C 14 C 17 The trust factor for each case is 0.8. Therefore, the case with the highest similarity to the target case is S′. 10 =T·S(C0,C 10 ) = 0.8 × 0.9 = 0.72, that is, C 10 As a candidate case, source case C is extracted. 10The sorting result L 10 This is the priority ranking result of the battlefield information.

[0118] Based on the above embodiments, as an optional implementation, if the decision-maker is not satisfied with the initial ranking results, the ranking results can be regenerated by collecting more cases and adjusting threshold parameters. After the operation concludes, the source case library needs to be updated promptly, mainly including two specific tasks: first, based on the ranking results and their corresponding ranking effects, new cases are written and added to the source case library; second, based on the decision-maker's evaluation feedback, the trust factors of the candidate cases are modified to update the original data of the candidate cases.

[0119] Adjustment of the trust factor. Let the target case of a certain joint operation be C0, and the final selected alternative cases be C... q The corresponding sorting result L q =(l q1 ,l q2 ,…,l qk ), where l qi Indicates alternative case C q Let i be the i-th ranking result proposed by the decision-maker, where i ∈ k, and k indicates that there are a total of k ranking results. Among these k ranking results, the decision-maker indicates that there are... This information is not needed; based on this, a sorting result L′ is further generated. q =(l′ q1 ,l′ q2 ,…,l′ qk′ ), where l′ qi′ Indicates alternative case C q The commander further proposes the i′th ranking result, where k′ represents the number of pieces of information reintroduced by the decision-maker. The success rate P of this ranking result can be expressed as:

[0120]

[0121] Assume this is candidate case C q The similarity value with the target case C0 is S 0q Source Case C q The initial trust factor is T′ q Then, based on the sorting results, the source case C... q The modified trust factor T q It can be represented as:

[0122]

[0123] Based on the priority ranking results of this battlefield information L 10Assuming a success rate of P = 0.7, then according to the above formula, the impact of this personalized recommendation practice on the source case C is... 10 Trust factor T 10 Make corrections:

[0124]

[0125] In summary, this invention addresses the problem of battlefield information prioritization. Based on the KSF method to create a source case library from historical battle examples or training data, it uses VSM to extract the source cases most similar to the current combat scenario. Then, it employs CBR's case similarity analysis method to generate the most similar candidate cases. The battlefield information prioritization results are then used as the current battlefield information prioritization results. The rationality and feasibility of the method are verified through examples. The C-KVC battlefield information prioritization method proposed in this paper can support scientific battlefield decision-making, improving the rationality and accuracy of battlefield information prioritization, and has certain theoretical and practical significance.

[0126] The present invention also provides a battlefield information priority sorting device, comprising:

[0127] The source case library construction module is used to filter and extract attributes and information from historical battlefield data, and to construct the source case library using the critical success factor method; the historical battlefield data includes historical battle data and training activity data.

[0128] The similar case set extraction module is used to extract a similar case set from the source case library based on the similarity between the situation analysis report of the target case and the situation analysis report of the source case using a vector space model;

[0129] The alternative case extraction module is used to select the alternative cases most similar to the target case from the set of similar cases based on the feature attribute set of the source case; the feature attribute set includes decision-maker style characteristics and combat activity characteristics;

[0130] The sorting result determination module is used to take the battlefield information priority sorting result of the candidate cases as the current battlefield information priority sorting result of the target case;

[0131] The source cases in the source case library are described using a quadruple representation; the quadruple representation is as follows:

[0132] C i ={R i ,F i ,L i ,T i |i∈N};

[0133] Among them, C iF represents the i-th source case in the source case set. i Let L represent the set of feature attributes corresponding to the i-th source case. i T represents the information priority ranking result corresponding to the i-th source case. i Let N represent the trust factor of the sorting result corresponding to the i-th source case, and N represent the number of source cases in the case library; and,

[0134] F i =(f ij |j=1,2,…,m);

[0135] Among them, f ij Let m represent the j-th feature attribute of the i-th source case, where m is the dimension of the feature attribute set; the feature attributes include decision-maker style features and operational activity features.

[0136] It should be noted that the battlefield information priority sorting device provided in this embodiment of the invention can execute the battlefield information priority sorting method described in any of the above embodiments during specific operation, which will not be elaborated in this embodiment.

[0137] Figure 6 This is a schematic diagram of the structure of the electronic device provided by the present invention, such as... Figure 6 As shown, the electronic device may include: a processor 610, a communications interface 620, a memory 630, and a communication bus 640, wherein the processor 610, communications interface 620, and memory 630 communicate with each other through the communication bus 640. The processor 610 can call logical instructions in the memory 630 to execute a battlefield information priority ranking method. This method includes: filtering and extracting attributes and information from historical battlefield data; constructing a source case library using the critical success factor method; the historical battlefield data includes historical battle data and training activity data; using a vector space model, extracting a set of similar cases from the source case library based on the similarity between the situation analysis report of the target case and the situation analysis report of the source case; selecting candidate cases most similar to the target case from the set of similar cases based on the characteristic attribute set of the source case; the characteristic attribute set includes decision-maker style characteristics and combat activity characteristics; and using the battlefield information priority ranking result of the candidate cases as the current battlefield information priority ranking result of the target case.

[0138] Furthermore, the logical instructions in the aforementioned memory 630 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0139] On the other hand, the present invention also provides a computer program product, the computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions, wherein when the program instructions are executed by a computer, the computer is able to execute the battlefield information priority ranking method provided in the above embodiments, the method comprising: filtering and extracting attributes and information from historical battlefield data, constructing a source case library using the critical success factor method; the historical battlefield data including historical battle data and training activity data; using a vector space model, extracting a set of similar cases from the source case library based on the similarity between the situation analysis report of the target case and the situation analysis report of the source case; filtering out candidate cases most similar to the target case from the set of similar cases based on the feature attribute set of the source case; the feature attribute set including decision-maker style characteristics and combat activity characteristics; and using the battlefield information priority ranking result of the candidate cases as the current battlefield information priority ranking result of the target case.

[0140] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements the battlefield information priority ranking method provided in the above embodiments. The method includes: filtering and extracting attributes and information from historical battlefield data; constructing a source case library using the critical success factor method; the historical battlefield data includes historical battle data and training activity data; extracting a set of similar cases from the source case library based on the similarity between the situation analysis report of the target case and the situation analysis report of the source case using a vector space model; selecting candidate cases most similar to the target case from the set of similar cases based on the feature attribute set of the source case; the feature attribute set includes decision-maker style characteristics and combat activity characteristics; and using the battlefield information priority ranking result of the candidate cases as the current battlefield information priority ranking result of the target case.

[0141] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0142] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0143] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for prioritizing battlefield information, characterized in that, include: Historical battlefield data is filtered and extracted for attributes and information, and a source case library is constructed using the critical success factor method; the historical battlefield data includes historical battle data and training activity data. Using the vector space model, a set of similar cases is extracted from the source case library based on the similarity between the situation analysis report of the target case and the situation analysis report of the source case. Based on the feature attribute set of the source case, select the candidate case that is most similar to the target case from the set of similar cases; The set of characteristic attributes includes decision-maker style characteristics and operational activity characteristics; The priority ranking result of the battlefield information of the candidate cases is used as the priority ranking result of the current battlefield information of the target case. The source cases in the source case library are described using a quadruple representation; the quadruple representation is as follows: C i ={R i ,F i ,L i ,T i |i∈N}; Among them, C i F represents the i-th source case in the source case set. i Let L represent the set of feature attributes corresponding to the i-th source case. i T represents the information priority ranking result corresponding to the i-th source case. i Let N represent the trust factor of the sorting result corresponding to the i-th source case, and N represent the number of source cases in the case library; and, F i =(f ij |j=1,2,…,m); Among them, f ij Let m represent the j-th feature attribute of the i-th source case, where m is the dimension of the feature attribute set; the feature attributes include decision-maker style features and operational activity features.

2. The battlefield information priority sorting method according to claim 1, characterized in that, After prioritizing the battlefield information of the candidate cases as the current battlefield information priority ranking result of the target case, the method further includes: Based on the ranking results and the decision-makers' feedback on the ranking, the source cases are modified and new source cases are generated to update the source case library.

3. The battlefield information priority sorting method according to claim 1, characterized in that, The method utilizes a vector space model to extract a set of similar cases from the source case database based on the similarity between the target case's situation analysis report and the source case's situation analysis report. This set includes: Obtain the target keyword vector from the situation analysis report of the target case, and the keyword vector from the situation analysis report of each source case; Based on each keyword vector and the target keyword vector, obtain the similarity between each source case and the target case; Based on the similarity between each source case and the target case, and a preset similarity threshold, a set of similar cases is extracted from the source case library.

4. The battlefield information priority sorting method according to claim 1, characterized in that, The step of selecting candidate cases most similar to the target case from the set of similar cases based on the feature attribute set of the source case includes: From the set of feature attributes, determine multiple target feature attributes that can better characterize the features of the source case; By using a combination of subjective and objective weighting methods, the combined weights of each target feature attribute are determined; Based on the type of the target feature attribute, different methods are used to calculate the similarity of each target feature attribute between each source case and the target case; The feature similarity between each source case and the target case is calculated based on the similarity of each target feature attribute between each source case and the target case, and the combined weight of each target feature attribute. Based on the trust factor and feature similarity of each source case, the source case most similar to the target case is determined as the candidate case.

5. The battlefield information priority sorting method according to claim 3, characterized in that, Before extracting a set of similar cases from the source case library based on the similarity between each source case and the target case and a preset similarity threshold, the process further includes: From the similarity between each source case and the target case, determine the maximum and minimum similarity. The preset similarity threshold is determined based on the maximum similarity and the minimum similarity.

6. The battlefield information priority sorting method according to claim 4, characterized in that, The method of combining subjective and objective weighting to determine the combined weight of each target feature attribute includes: Determine the subjective weight vector and objective weight vector of the target feature attributes; The target weight vector of the target feature attribute is determined based on the standard deviation of the subjective weight vector, the standard deviation of the objective weight vector, the subjective weight vector, and the objective weight vector.

7. The battlefield information priority sorting method according to claim 2, characterized in that, The step of modifying source cases and generating new source cases based on the ranking effect and the decision-maker's evaluation feedback to update the source case library includes: Based on the sorting results and the corresponding sorting effect, write new cases and add them to the source case library; Based on the evaluation feedback from decision-makers, the trust factor of the alternative cases is modified to update the original data of the alternative cases.

8. A battlefield information priority sorting device, characterized in that, include: The source case library construction module is used to filter and extract attributes and information from historical battlefield data, and to construct the source case library using the critical success factor method; the historical battlefield data includes historical battle data and training activity data. The similar case set extraction module is used to extract a similar case set from the source case library based on the similarity between the situation analysis report of the target case and the situation analysis report of the source case using a vector space model; The alternative case extraction module is used to select the alternative cases that are most similar to the target case from the set of similar cases based on the feature attribute set of the source case; The set of characteristic attributes includes decision-maker style characteristics and operational activity characteristics; The ranking result determination module is used to take the battlefield information priority ranking result of the candidate cases as the current battlefield information priority ranking result of the target case. The source cases in the source case library are described using a quadruple representation; the quadruple representation is as follows: C i ={R i ,F i ,L i ,T i |i∈N}; Among them, C i F represents the i-th source case in the source case set. i Let L represent the set of feature attributes corresponding to the i-th source case. i T represents the information priority ranking result corresponding to the i-th source case. i Let N represent the trust factor of the sorting result corresponding to the i-th source case, and N represent the number of source cases in the case library; and, F i =(f ij |j=1,2,…,m); Among them, f ij Let m represent the j-th feature attribute of the i-th source case, where m is the dimension of the feature attribute set; the feature attributes include decision-maker style features and operational activity features.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the battlefield information priority sorting method as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the battlefield information priority sorting method as described in any one of claims 1 to 7.

Images