A method and device for analyzing electronic case content in a smart court

Abstract

The present application relates to the field of natural language processing technology, and specifically to a method and device for analyzing the content of electronic case files in a smart court, which specifically includes: obtaining the order difference measurement of the same keywords between different electronic case files based on the differences between sentences containing the same keywords in different electronic case files, and constructing the measurement coefficient of each keyword in each electronic case file in combination with the TF-IDF value and part of speech of each keyword in each electronic case file; constructing a weight function according to the measurement coefficient of the keyword, and calculating the weight of each keyword; clustering according to the fingerprint of the electronic case file and the keyword measurement coefficient, and obtaining the core keyword set of each cluster. It improves the accuracy of the division of cases of the same type and the extraction of keywords, improves the accuracy of the measurement of the similarity of electronic case files, and helps to improve the accuracy of the content analysis of electronic case files and the division of case types.

Description

Technical Field

[0001] The present application relates to the field of natural language processing technology, and specifically to a method and device for analyzing the content of electronic case files in a smart court. Background Art

[0002] With the acceleration of digital transformation, the court system is also improving in the direction of intelligence and digitization to improve the efficiency and transparency of judicial work. As an important part of the construction of smart courts, electronic case files are huge in number and complex in content. Therefore, an efficient electronic case file content analysis method is crucial to improving judicial efficiency and optimizing the storage of electronic case files. The existing electronic case file storage can choose to store cases of the same type together, so that the retrieval process can be realized quickly and efficiently, which requires accurate distinction between electronic case files of the same type of cases.

[0003] When analyzing the content of electronic documents, the prior art will calculate the text similarity and distinguish them according to the similarity of the electronic documents. The SimHash algorithm is a commonly used text similarity calculation method. The algorithm will generate a fingerprint for each electronic document. Since the generated fingerprint has good local sensitivity and high retrieval efficiency, it is widely used in similarity detection of large amounts of text. However, when processing electronic documents, the SimHash algorithm only considers keywords based on the word frequency dimension, ignoring the impact of other keyword features on the electronic documents. Electronic documents are a type of text with strong logic. Relying only on word frequency will ignore the logical order characteristics of keywords, resulting in inaccurate classification of similar electronic documents and extraction of core keywords. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of this application is to provide a method and device for analyzing the content of electronic case files in a smart court. The technical solutions adopted are as follows:

[0005] In a first aspect, an embodiment of the present application provides a method for analyzing electronic case contents of a smart court, the method comprising the following steps:

[0006] Segment the text of each electronic case to obtain each phrase, and remove stop words from all phrases to obtain each keyword;

[0007] Based on the differences between sentences containing the same keywords in different electronic files, construct the order measurement coefficients of the same keywords between the sentences in different electronic files; based on the differences in the order measurement coefficients between the sentences in different electronic files, obtain the order difference measurement of the same keywords between different electronic files; based on the importance and part-of-speech features of each keyword in each electronic file, combined with the order difference measurement, construct the measurement coefficient of each keyword in each electronic file;

[0008] Determine the weight of each keyword in each electronic case based on the measurement coefficient, and obtain the fingerprint of each electronic case in combination with the fingerprint generation algorithm;

[0009] Clustering is performed based on the maximum metric coefficient and fingerprint in each electronic case to obtain each cluster;

[0010] Based on the keywords of each electronic case in each cluster, a core keyword set of each cluster is obtained.

[0011] In one embodiment, the calculation expression of the order metric coefficient is:

[0012] , where It is a statement With statement The order metric coefficient of keyword i between ; , The statements are and statements The ordinal value of keyword i in , , The statements are and statements The length of the sentence, is a preset positive number, is an exponential function with the natural constant e as base;

[0013] The statement is the sentence containing keyword i in the zth sentence in the Fth electronic case. is the tth sentence containing keyword i in the Sth electronic case.

[0014] In one embodiment, the calculation expression of the order difference metric is:

[0015] ,in, is the order difference measure of keyword i between the Fth electronic case and the Sth electronic case, It is a statement With statement The order metric coefficient of keyword i between , are the number of sentences containing keyword i in the Fth electronic case and the Sth electronic case, is the minimum value function.

[0016] In one embodiment, the process of obtaining the measurement coefficient is as follows:

[0017] For any keyword in any electronic case, the average of the order difference measures of the keyword between the electronic case and all other electronic cases is used as the comprehensive order difference measure of the keyword in the electronic case;

[0018] Obtain the TF-IDF value of each keyword in each electronic case, and set the part-of-speech metric of each keyword based on the part-of-speech of each keyword;

[0019] In any electronic case, the measurement coefficient of each keyword is positively correlated with the TF-IDF value and part-of-speech measurement of each keyword, and negatively correlated with the comprehensive order difference measurement of each keyword.

[0020] In one embodiment, the part-of-speech measure of each keyword is: the part-of-speech measure of the keyword of the verb part-of-speech is set to 1, and the part-of-speech measures of the keywords of other parts of speech are set to 0.

[0021] In one embodiment, the calculation expression of the weight of each keyword in each electronic case is:

[0022] , where is the weight of keyword i in the Fth electronic case, is the metric coefficient of keyword i in the Fth electronic case, is the number of keywords in the Fth electronic case, To find the maximum value function.

[0023] In one embodiment, the process of obtaining the fingerprint of each electronic case file is as follows:

[0024] The sequence composed of all keywords in each electronic file and the sequence composed of the weights of all keywords are used as the input of the fingerprint generation algorithm, and the output is the fingerprint of each electronic file.

[0025] In one embodiment, the process of obtaining each cluster is as follows:

[0026] The maximum value of all keyword measurement coefficients in each electronic case is converted into a binary number, and the fingerprint of each electronic case and the binary number are combined into a vector; the vector of all electronic cases is used as the input of the clustering algorithm, and the output is each cluster.

[0027] In one embodiment, the process of obtaining the core keyword set is as follows:

[0028] In each electronic case, a set of keywords with a metric coefficient greater than a preset segmentation threshold is recorded as a key keyword set;

[0029] In each cluster, the intersection of the key keyword sets of all electronic files and all verbs are combined to form a set as the core keyword set of each cluster.

[0030] In the second aspect, an embodiment of the present application also provides an electronic case content analysis device for a smart court, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above-mentioned methods when executing the computer program.

[0031] The embodiments of the present application have at least the following beneficial effects:

[0032] This application obtains the order difference measurement of the same keywords between different electronic case files based on the differences between sentences containing the same keywords in different electronic case files, and constructs the measurement coefficient of each keyword in each electronic case file in combination with the TF-IDF value and part of speech of each keyword in each electronic case file. This indicator takes into account the frequency and logical order of the keywords at the same time, which helps to improve the accuracy of judging the importance of keywords; constructs a weight function according to the measurement coefficient of the keyword, and calculates the weight of each keyword, which helps to improve the accuracy of calculating the electronic case file fingerprint using the SimHash algorithm, and improves the accuracy of subsequent electronic case file similarity judgment; clusters the electronic case files according to the fingerprints and keyword measurement coefficients, and obtains the core keyword set of each cluster, which improves the accuracy of the division of cases of the same type and the extraction of keywords, improves the accuracy of measuring the similarity of electronic cases, and helps to improve the accuracy of electronic case file content analysis and case type division. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0034] Figure 1 A flowchart of a method for analyzing electronic case contents of a smart court provided in accordance with an embodiment of the present application;

[0035] Figure 2 Schematic diagram of the process of obtaining the measurement coefficient of each keyword in each electronic case. DETAILED DESCRIPTION

[0036] In order to further explain the technical means and effects adopted by this application to achieve the predetermined invention purpose, the following is a detailed description of the electronic case content analysis method and device of a smart court proposed in this application, its specific implementation method, structure, features and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0037] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0038] The following is a detailed description of the specific scheme of the electronic case content analysis method and device for a smart court provided by the present application in conjunction with the accompanying drawings.

[0039] See also Figure 1 , which shows a flowchart of a method for analyzing electronic case contents of a smart court provided by an embodiment of the present application, the method comprising the following steps:

[0040] Step S1, segment the text content of each electronic case to obtain each phrase, and remove stop words from all phrases to obtain each keyword.

[0041] In addition to the main text, the electronic case also includes a title and an ending signature. The title and the ending signature are redundant information. The title and the ending signature of each electronic case are filtered out through rule matching, thereby obtaining the main text data of each electronic case. The LTP toolkit can process the content in the document by sentence segmentation, word segmentation, part-of-speech tagging, and stop word deletion. Using the LTP toolkit to divide each electronic case into sentences, a sentence set can be obtained; each sentence is segmented separately to obtain each phrase, and each phrase in each sentence is tagged with the part of speech; the Harbin Institute of Technology stop word list is used to remove stop words from all phrases after the part-of-speech tagging, and stop words such as conjunctions, prepositions, and articles are deleted. The remaining phrases are called keywords. Among them, rule matching and LTP toolkit are well-known technologies, and the specific process will not be repeated.

[0042] Step S2, constructing a measurement coefficient for each keyword in each electronic case file according to the differences between sentences containing the same keyword in different electronic cases, the TF-IDF value and the part of speech of each keyword in each electronic case file.

[0043] Electronic case files are a kind of highly logical text. Due to the particularity of their usage scenarios, they do not allow for overly personalized expressions. For the same type of case, the expressions of events and related content in the electronic case files are highly similar. However, since everyone's expression methods are different, this high degree of similarity is mainly reflected in the relevant keywords. Therefore, the keyword features of electronic case files can be used to calculate the similarity between different electronic case files, and representative keywords of the same type of case can be extracted.

[0044] Cases of the same type often contain many similar statements. At the same time, due to the textual nature of case files, some sentences that can characterize the type of case have the same keywords, and the order in which the keywords appear is fixed. Since the event described in the electronic case file is an action imposed by one individual on another individual, and the action is one of the important information in the electronic case file, it will be mentioned many times in the electronic case file. Therefore, when a keyword appears more frequently in an electronic case file and the keyword is a verb, the information contained in the keyword is more important, and it is also more important for the electronic case file. However, for all electronic case files, the more frequently a keyword appears in all case files, the lower the contribution rate of the keyword to distinguishing different types of cases.

[0045] Based on the above analysis, (1) according to the differences between sentences containing the same keyword in different electronic documents, the order measurement coefficient between each sentence containing any keyword in any electronic document and each sentence containing the same keyword in each remaining electronic document is constructed to measure the logical order of each keyword:

[0046] In an electronic case, a keyword may be contained in multiple sentences. The sentence containing keyword i in the zth sentence of the Fth electronic case is recorded as , record the tth sentence containing keyword i in the Sth electronic case as , statement With statement The calculation expression of the order measurement coefficient of the keyword i between is:

[0047] , where It is a statement With statement The order metric coefficient of keyword i between ; , The statements are and statements The ordinal value of keyword i in , , The statements are and statements The length of the sentence, is an exponential function with natural number e as base, is a preset positive number, which is used to prevent the denominator from being 0. Preferably, in one embodiment of the present application, The value of is set to 1. In other embodiments, the implementer can set it according to the actual situation. The value of .

[0048] When a keyword is a keyword in the basis for judging the case type in an electronic case file, the statement of the basis should exist in any two electronic case files of the same type of case, that is, in the two electronic case files, there is at least one sentence in which the keyword appears in the same or close order, and the length of the phrase sequence of the sentence is also the same or close, that is, the numerator is close to 1, and the denominator is also close to 1; therefore, when two electronic case files correspond to the same type of case, and a keyword is a keyword in the basis for judging this type of case, the order measurement coefficient of the keyword is approximately 1. On the contrary, when a keyword is not a keyword in the basis for judging a type of case, its order measurement coefficient is much less than 1 or much greater than 1, and the farther the distance from 1, the smaller the contribution of this keyword to distinguishing case types and the less important it is.

[0049] (2) Obtain the order difference measurement of the same keywords between different electronic files based on the order measurement coefficient:

[0050] ,in, is the order difference measure of keyword i between the Fth electronic case and the Sth electronic case, It is a statement With statement The order metric coefficient of keyword i between , are the number of sentences containing keyword i in the Fth electronic case and the Sth electronic case, is the minimum value function;

[0051] As the order measurement coefficient is closer to 1, the corresponding keyword is more important, and the order difference measurement is closer to 0.

[0052] (3) The mean of the order difference measures of keyword i between the Fth electronic case and all other electronic cases is taken as the comprehensive order difference measure of keyword i in the Fth electronic case.

[0053] (4) Calculate the TF-IDF value (term frequency inverse document frequency) of each keyword in each electronic case file, where the TF-IDF value is a well-known technology and the specific process is not repeated here; preset the part-of-speech measure of each keyword according to the part-of-speech of each keyword. Since the information contained in the verb in the electronic case file is more important, the part-of-speech measure of the keyword with the verb part-of-speech is set to 1, and the part-of-speech measure of the keyword with the other parts of speech is set to 0, so that the higher the part-of-speech measure of the keyword, the more important the keyword. It should be noted that the implementer of the part-of-speech measure of the keyword can set it by himself, and the embodiment of the present application does not make any specific restrictions.

[0054] (5) In order to measure the contribution of each keyword to distinguishing case types or describing case events, in any electronic case file, a measurement coefficient for each keyword in each electronic case file is constructed based on the TF-IDF value, comprehensive order difference measure and part-of-speech measure of each keyword:

[0055] In any electronic case, the measurement coefficient of each keyword is positively correlated with the TF-IDF value and part-of-speech measurement of each keyword, and negatively correlated with the comprehensive order difference measurement of each keyword.

[0056] It can be understood that the positive correlation and negative correlation in this application refer to the relationship between the independent variable and the dependent variable. The positive correlation means that the independent variable increases (decreases) as the dependent variable increases (decreases); the negative correlation means that the independent variable decreases (increases) as the dependent variable increases (decreases); the specific positive correlation and negative correlation can be determined according to actual conditions during the application process, and this application does not impose any special restrictions.

[0057] Preferably, in one embodiment of the present application, the calculation expression of the measurement coefficient of each keyword in any electronic case file can be: , where is the metric coefficient of keyword i in the Fth electronic case, is the comprehensive order difference measure of keyword i in the Fth electronic case, is the TF-IDF value of keyword i in the Fth electronic case, is the part-of-speech measure of keyword i in the Fth electronic case, is a preset positive number, which is used to avoid the denominator being 0. Preferably, in one embodiment of the present application, The value of is set to 1. In other embodiments of the present application, the implementer can set it according to the actual situation. The value of .

[0058] It is understandable that when a keyword contributes more to distinguishing case types or describing case events, the logical order of the keyword is relatively fixed, that is, the order in which the keyword appears in a sentence and the length of the sentence are relatively fixed or approximate; secondly, the keyword's word frequency inverse document frequency is greater, and if the keyword is a verb, it may be a description of the event behavior, which is relatively more important; thus, when a keyword contributes more to distinguishing case types or describing case events, the metric coefficient of the keyword is larger, and the larger the metric coefficient, the more important the keyword. Conversely, when the keyword contributes less to distinguishing case types or describing case events, the metric coefficient of the keyword is smaller.

[0059] Step S3, calculating the weight of each keyword in each electronic case file according to the measurement coefficient of each keyword in each electronic case file, and generating the fingerprint of each electronic case file in combination with the SimHash algorithm.

[0060] (1) In each electronic case, the greater the contribution of a keyword, that is, the greater the measurement coefficient, the more effective information the keyword can provide for distinguishing case types or describing case events. Therefore, the more important the keyword is, the greater the weight should be given, which has a greater impact on the subsequent generation of electronic case fingerprints and the judgment of similarity, and helps to improve the accuracy of distinguishing case types based on electronic case files. The calculation expression for the weight of each keyword in any electronic case is:

[0061] , where is the weight of keyword i in the Fth electronic case, is the metric coefficient of keyword i in the Fth electronic case, is the number of keywords in the Fth electronic case, To find the maximum value function.

[0062] The greater the contribution of a keyword to distinguishing case types or describing case events, the greater the keyword's measurement coefficient, and thus the corresponding weight. The role of the denominator is to limit the value range of the weight to between 0 and 1.

[0063] (2) In each electronic case file, all keywords of all sentences are arranged in the order of their appearance in the electronic case file to obtain the phrase sequence of the electronic case file; then the weights of all keywords are arranged in the order of the keywords in the phrase sequence of the electronic case file to form the keyword weight sequence of the electronic case file; then the phrase sequence and keyword weight sequence of each electronic case file are used as the input of the SimHash algorithm, and the output is the fingerprint of each electronic case file. The SimHash algorithm is a well-known technology, and the specific process is not repeated here.

[0064] Step S4, clustering is performed based on the fingerprint of each electronic case and the maximum metric coefficient in each electronic case to obtain each cluster.

[0065] The maximum value of the measurement coefficients of all keywords in each electronic case is obtained and recorded as the maximum measurement coefficient; the maximum measurement coefficient is converted into a binary number. The vector composed of the fingerprint of each electronic case and the binary number of the maximum measurement coefficient of each electronic case is used as the feature vector of each electronic case.

[0066] The characteristic vectors of all electronic case files are used as the input of the K-means clustering algorithm, and the number of cluster centers is set. Preferably, in the embodiment of the present application, the number of cluster centers is set to 5. In other embodiments of the present application, the implementer can set the number of cluster centers according to the actual situation. Since the fingerprint and the metric coefficient are both in the form of strings, the distance measurement method used is the Hamming distance, and the output is each cluster cluster. Among them, the K-means algorithm is a well-known technology, and the specific process will not be repeated. It can be understood that for the clustering of the characteristic vectors of all electronic case files, this application only provides a clustering method. There are many existing clustering methods. The implementer can also use other clustering methods for clustering, and this application does not make specific restrictions.

[0067] The role of the maximum measurement coefficient in the electronic case is that if it is an electronic case of the same type of case, the description of the event in the electronic case is relatively similar, and the description of the basis is basically the same, so the measurement coefficients of the keywords should be basically equal. However, since the number of keywords in the electronic case is not necessarily the same, it is impossible to compare each measurement coefficient one by one, so the maximum measurement coefficient is used as a representative for clustering.

[0068] Step S5, obtaining a core keyword set of each cluster based on the keywords of each electronic case in each cluster.

[0069] In each electronic case, the quartile of the measurement coefficients of all keywords is used as the preset segmentation threshold of the electronic case, and the set of all keywords whose measurement coefficients are greater than the preset segmentation threshold is recorded as the key keyword set of the electronic case; it should be noted that, for the selection of the preset segmentation threshold, this application only provides a preset segmentation threshold selection method, and there are many existing preset segmentation threshold acquisition methods, and implementers may also use other methods to preset the segmentation threshold.

[0070] Obtaining the intersection of key keyword sets of all electronic cases in each cluster, and using the intersection as the key keyword set of the cluster;

[0071] The set of verbs in all electronic cases in the cluster is taken as the verb set of the cluster, wherein each element in the verb set does not appear repeatedly;

[0072] The union of the key keyword set and verb set of the cluster is taken as the core keyword set of the cluster. Each cluster represents the electronic case file of a type of case, and the core keyword set of a cluster represents the keyword set that contributes most to the description of this type of case, that is, the keywords in the core keyword set contribute most to distinguishing case types and describing case events.

[0073] In this way, when using the SimHash algorithm to calculate text similarity, not only the frequency of each keyword in the text is taken into account, but also the logical order of each keyword in the sentence, so that the calculation and judgment of the similarity of electronic case files are more accurate, and then the electronic case files of the same type of cases can be more accurately divided into the same cluster; secondly, by calculating the core keyword set of each cluster, it is possible to select the keywords that contribute most to the case type and event description, which helps to improve the extraction of key keywords for each type of case and improve the description of each type of case, thereby achieving accurate analysis of the content of electronic case files and precise division of case types.

[0074] The schematic diagram of the process of obtaining the measurement coefficient of each keyword in each electronic case is as follows: Figure 2 shown.

[0075] Based on the same inventive concept as the above method, an embodiment of the present application also provides an electronic case content analysis device for a smart court, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above-mentioned methods for analyzing the content of electronic case files for a smart court when executing the computer program.

[0076] In summary, the embodiment of the present application provides a method for analyzing the content of electronic case files of a smart court. The present application obtains the order difference measurement of the same keywords between different electronic case files based on the differences between sentences containing the same keywords in different electronic case files, and constructs the measurement coefficient of each keyword in each electronic case file in combination with the TF-IDF value and part of speech of each keyword in each electronic case file. This indicator takes into account the frequency and logical order of the keywords at the same time, which helps to improve the accuracy of judging the importance of keywords; constructs a weight function according to the measurement coefficient of the keyword, and calculates the weight of each keyword, which helps to improve the accuracy of calculating the fingerprint of the electronic case file using the SimHash algorithm, and improves the accuracy of subsequent similarity judgment of the electronic case file; finally, clustering is performed according to the fingerprint of the electronic case file and the keyword measurement coefficient, and the core keyword set of each cluster cluster is obtained, which improves the accuracy of the division of cases of the same type and the extraction of keywords, improves the accuracy of measuring the similarity of electronic case files, and helps to improve the accuracy of content analysis of electronic case files and classification of case types.

[0077] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. The above-mentioned specific embodiments of the present application are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0078] The various embodiments in the present application are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0079] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present application should be included in the protection scope of the present application.

Claims

1. A method for analyzing the content of electronic case files in a smart court, characterized in that: The method comprises the following steps: Segment the text of each electronic case to obtain each phrase, and remove stop words from all phrases to obtain each keyword; Based on the differences between sentences containing the same keywords in different electronic files, construct the order measurement coefficients of the same keywords between the sentences in different electronic files; based on the differences in the order measurement coefficients between the sentences in different electronic files, obtain the order difference measurement of the same keywords between different electronic files; based on the importance and part-of-speech features of each keyword in each electronic file, combined with the order difference measurement, construct the measurement coefficient of each keyword in each electronic file; Determine the weight of each keyword in each electronic case based on the measurement coefficient, and obtain the fingerprint of each electronic case in combination with the fingerprint generation algorithm; Clustering is performed based on the maximum metric coefficient and fingerprint in each electronic case to obtain each cluster; Obtaining a core keyword set of each cluster based on the keywords of each electronic case in each cluster; The calculation expression of the order measurement coefficient is: , where It is a statement With statement The order metric coefficient of keyword i between ; , The statements are and statements The ordinal value of keyword i in , , The statements are and statements The length of the sentence, is a preset positive number, is an exponential function with the natural constant e as base; The statement is the sentence containing keyword i in the zth sentence in the Fth electronic case. is the tth sentence containing keyword i in the Sth electronic case; The calculation expression of the order difference metric is: ,in, is the order difference measure of keyword i between the Fth electronic case and the Sth electronic case, It is a statement With statement The order metric coefficient of keyword i between , are the number of sentences containing keyword i in the Fth electronic case and the Sth electronic case, is the minimum value function; The process of obtaining the core keyword set is as follows: In each electronic case, a set of keywords with a metric coefficient greater than a preset segmentation threshold is recorded as a key keyword set; In each cluster, the intersection of the key keyword set of all electronic files and the verbs of all electronic files are combined to form a set as the core keyword set of each cluster; The process of obtaining the measurement coefficient is as follows: For any keyword in any electronic case, the average of the order difference measures of the keyword between the electronic case and all other electronic cases is used as the comprehensive order difference measure of the keyword in the electronic case; Obtain the TF-IDF value of each keyword in each electronic case, and set the part-of-speech metric of each keyword based on the part-of-speech of each keyword; In any electronic case, the metric coefficient of each keyword is positively correlated with the TF-IDF value and part-of-speech metric of each keyword, and negatively correlated with the comprehensive order difference metric of each keyword; The calculation expression of the weight of each keyword in each electronic case is: , where is the weight of keyword i in the Fth electronic case, is the metric coefficient of keyword i in the Fth electronic case, is the number of keywords in the Fth electronic case, To find the maximum value function; The process of obtaining the fingerprint of each electronic case is as follows: The sequence of all keywords in each electronic case file and the sequence of all keyword weights are used as the input of the fingerprint generation algorithm, and the output is the fingerprint of each electronic case file; The process of obtaining each cluster is as follows: The maximum value of all keyword measurement coefficients in each electronic case is converted into a binary number, and the fingerprint of each electronic case and the binary number are combined into a vector; the vector of all electronic cases is used as the input of the clustering algorithm, and the output is each cluster.

2. The method for analyzing electronic case contents of a smart court as claimed in claim 1, characterized in that: The part-of-speech measure of each keyword is: the part-of-speech measure of the keyword with the verb part-of-speech is set to 1, and the part-of-speech measures of the keywords with other parts of speech are set to 0.

3. A device for analyzing electronic case contents of a smart court, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 2 are implemented.

Images