Video conference equipment fault diagnosis method based on knowledge base

Abstract

The invention relates to the field of equipment fault diagnosis methods, in particular to a video conference equipment fault diagnosis method based on a knowledge base. According to the method, the segmented words in the first data are extracted through the first data containing a plurality of historical fault logs, and then the plurality of segmented words are screened out from all aspects as dimensions by gradually calculating the first importance degree, the second importance degree and the third importance degree, so that the plurality of feature words relative to the first data are formed; then, when third data is obtained, if first data identical to the third data exists, corresponding second data is directly output as a processing method, and once the second data cannot solve faults recorded in the third data, the second data for reference is provided for operation and maintenance personnel through final fitting degree calculation; the problem that at present, when operation and maintenance personnel manually diagnose video conference equipment faults step by step, the requirement for rapidly diagnosing the equipment faults in the conference process is difficult to meet is solved.

Description

Technical Field

[0001] The invention relates to the field of equipment fault diagnosis methods, and in particular to a video conference equipment fault diagnosis method based on a knowledge base. Background Art

[0002] In large video conference rooms, due to the large number of devices, automatic monitoring of device status is currently achieved by setting logs, so that operation and maintenance personnel can timely understand the operating status of each device, such as current, voltage, network and other information. However, due to the large number of devices, there are relatively more faults. After each fault occurs, most of them require operation and maintenance personnel to conduct detection and step-by-step inspection, so as to finally identify and diagnose the cause of the fault and repair it. At present, only for some faults with a high probability of occurrence, operation and maintenance personnel can quickly solve the cause of the fault without step-by-step diagnosis or only need a small amount of diagnosis. Therefore, in daily use, since large conference rooms are often used for important meetings, the relatively slow step-by-step diagnosis method of operation and maintenance personnel can hardly meet the needs of rapid diagnosis of faults during meetings. Summary of the invention

[0003] In view of the deficiencies in the prior art, the present invention provides a video conferencing equipment fault diagnosis method based on a knowledge base, which solves the problem that it is difficult to meet the demand for rapid diagnosis of equipment faults during a meeting when the operation and maintenance personnel currently diagnose video conferencing equipment faults step by step manually.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] A video conferencing equipment fault diagnosis method based on a knowledge base, the diagnosis method comprising the following steps:

[0006] S1, obtaining a plurality of first data including historical fault logs of video conferencing equipment, and second data of a method used to solve the fault corresponding to the first data;

[0007] S2, calculating the total importance of each word in the first data, extracting several words with the highest total importance as feature words, and establishing a knowledge base of the corresponding relationship between the first data and the second data based on the feature words;

[0008] S3, obtaining third data including the current fault log, and determining whether there is first data identical to the third data;

[0009] If yes, then output the second data corresponding to the first data according to the knowledge base and proceed to step S4;

[0010] If not, proceed to step S5;

[0011] S4, determining whether the second data has completed processing the third data;

[0012] If so, then end;

[0013] If not, proceed to step S5;

[0014] S5. Obtain feature words of the third data, convert the feature words in the first data and the third data into vectors, calculate the degree of fit between the third data and each first data, and output the second data corresponding to a number of first data with the greatest degree of fit.

[0015] Preferably, in step S2, the following steps are specifically included:

[0016] S21, setting a custom word library according to the first data, and performing word segmentation extraction on the first data respectively to obtain a plurality of word segmentation libraries, and setting part-of-speech tags for the extracted word segmentations;

[0017] S22, setting a first window, traversing each word in each word library in turn, and setting connecting edges between the words that appear in the same first window at the same time to obtain a connecting edge set;

[0018] S23, obtaining a number of word segments with the largest number of connection edges according to the connection edge set to obtain a key word set;

[0019] S24, calculating the frequency importance and overall importance of each word in the key word set, and calculating the first importance according to the product of the two;

[0020] S25, calculating the confusion degree of any word in the key word set according to the probability of occurrence of the word in the corresponding word library and the total number of occurrences of all words in the word library, and calculating the second importance according to the part of speech of the word and the confusion degree;

[0021] S26, calculating the third importance of any participle according to its position in the key word set;

[0022] S27, calculating the total importance of any word in the key word set according to the first importance, the second importance, and the third importance;

[0023] S28. Obtain several word segments with the largest total importance values ​​in each word segment library in turn, mark them as feature words, and correspond them to the second data to construct a knowledge base.

[0024] Preferably, in step S24, the following steps are specifically included:

[0025] S241, sequentially obtain the number of occurrences of any word in the key word set in the corresponding word library and the total number of occurrences of all words in the word library, and calculate the ratio of the two to obtain the frequency importance of the word; the calculation formula of the frequency importance is:

[0026]

[0027] In the above formula, Z p Indicates the frequency importance of the word. Indicates that the i-th word is in the corresponding word library m i The number of times it appears in Indicates the number of times the jth word appears in the word library. Represents the word segmentation library m i The total number of occurrences of all participles in ;

[0028] S242, calculating the distinction importance of a segmented word according to the total number of segmented word libraries and whether each segmented word library contains any segmented word, and calculating the distinction importance of each segmented word in the key word set in turn;

[0029] S243, calculating the first importance according to the product of the word frequency importance and the distinction importance;

[0030] Z1=Z p ×Z q

[0031] In the above formula, Z1 represents the first importance, Z p Indicates the frequency importance of the word, Z q Indicates the importance of distinction.

[0032] Preferably, in step S25, the following steps are specifically included:

[0033] S251. Calculate the chaos degree of any word in the key word set according to the probability of occurrence of the word in the corresponding word library and the total number of occurrences of all words in the word library. The calculation formula of the chaos degree is:

[0034]

[0035] In the above formula, H(c i ) indicates the focus word concentration participle c i The degree of disorder, n m Indicates the total number of occurrences of all words in all word libraries. Indicates the focus word concentration participle c i The total number of occurrences in all word segments;

[0036] S252: Calculate the second importance of each word in the key word set according to the confusion degree and the part of speech of each word in the key word set.

[0037] Preferably, in step S26, the following steps are specifically included:

[0038] S261, constructing a connection matrix of the key word set; the expression of the connection matrix is:

[0039]

[0040] In the above formula, M Q Represents the connection matrix of the key word set, G k,k-1 Indicates whether there is a connecting edge between the kth word and the k-1th word in the key word set. If there is a connecting edge, the result is 1, otherwise it is 0;

[0041] S262, calculating the third importance of any word in the key word set according to the connection matrix;

[0042]

[0043] In the above formula, Z3 represents the third importance of any word in the key word set. and Respectively represent the positions of the i-th and j-th words in the key word set, and the results are i and j respectively, where is the position of the word with the largest number of connecting edges.

[0044] Preferably, in step S5, the following steps are specifically included:

[0045] S51, obtaining feature words of the third data, and converting the feature words in the first data and the third data into vectors;

[0046] S52, calculating the fit between the first data and the third data according to the vector of the feature word;

[0047] S53, extracting a plurality of first data with the greatest fit, and outputting the corresponding second data.

[0048] Preferably, in step S242, the calculation formula of the difference importance is:

[0049]

[0050] In the above formula, Z q Indicates the importance of each word in the key word set, n m Indicates the total number of word libraries, count(c j ∈m) means that the statistics contain word c j The total number of word libraries.

[0051] Preferably, in step S252, the calculation formula of the second importance is:

[0052] Z2=γ×H(c i )

[0053] in,

[0054]

[0055] In the above formula, Z2 is the second importance, γ is the part-of-speech coefficient, γ1, γ2 and γ3 are the first part-of-speech coefficient, the second part-of-speech coefficient and the third part-of-speech coefficient respectively.

[0056] Preferably, in step S252, γ1, γ2 and γ3 are set to 0.5, 0.4 and 0.1 respectively.

[0057] Preferably, in step S52, the calculation formula of the fit degree is:

[0058]

[0059] In the above formula, TH represents the fit between the first data and the third data, I ik and I jk Represents the vector of the kth feature word in the first data and the third data respectively. There are P feature words in total. T i is a frequency weight coefficient, and sets the output frequency of the second data corresponding to the first data. The frequency weight coefficient is positively correlated with the output frequency.

[0060] Compared with the prior art, the present invention provides a method for diagnosing video conferencing equipment faults based on a knowledge base, which has the following beneficial effects:

[0061] 1. The present invention first extracts the segmented words in the first data including several historical fault logs, and then gradually calculates the first importance, the second importance and the third importance, and selects several segmented words from various aspects as dimensions to form several feature words relative to the first data. Then, when the third data is obtained, if there is the same first data as the third data, the corresponding second data is directly output as a processing method. Once the second data cannot solve the fault recorded in the third data, at this time, the final fit calculation is used to provide the operation and maintenance personnel with the second data for reference, so as to assist the operation and maintenance personnel in quickly solving the fault and ensure the normal progress of the video conference.

[0062] 2. The present invention calculates the word frequency importance and the overall importance to select the word segments that are different from other first data in the key word set corresponding to the first data, thereby increasing the distinctiveness between the first data. On the one hand, the first data similar to the third data can be selected more accurately in the subsequent process to more accurately obtain the accurate second data.

[0063] 3. By calculating the second importance according to the part of speech of the participle, the participle can be classified according to the part of speech, so that some unimportant participles can be eliminated to extract the important participles in the first data and form feature words that are more corresponding to the first data.

[0064] 4. The present invention calculates the third importance according to the position of the segmented words in the key word set, so that it can form feature words that are more corresponding to the first data according to the relative position differences between different segmented words, thereby making the acquired feature words more corresponding to the first data. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0066] Figure 1 A flow chart of a method for diagnosing video conferencing equipment faults based on a knowledge base of the present invention;

[0067] Figure 2 This is the actual operation interface diagram of the knowledge base. DETAILED DESCRIPTION

[0068] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific implementation methods, so that the implementation process of how the present application uses technical means to solve technical problems and achieve technical effects can be fully understood and implemented accordingly.

[0069] Those skilled in the art can understand that all or part of the steps in the following embodiments can be completed by instructing the relevant hardware through a program, so the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0070] To solve the problem that it is difficult to meet the requirement of quickly diagnosing equipment failures during a meeting when maintaining personnel manually diagnose video conferencing equipment failures step by step at present, the present invention provides a method for diagnosing video conferencing equipment failures based on a knowledge base. In the terminal control system, the main equipment in the meeting room is docked. When a failure occurs in these main equipment, the system will synchronously receive the failure information log. By recording the occurred failures and failure handling methods in advance for the equipment in the system, when the equipment gives an alarm, the information of the failure handling methods entered in the knowledge base will be synchronously sent out for reminder. During failure handling, technicians refer to a certain processing suggestion in the knowledge base to handle the failure, and correspond the final processing method with the failure information log. The following further details this diagnostic method. This diagnostic method includes the following steps:

[0071] S1. Obtain a number of first data containing historical failure logs of video conferencing equipment, and second data on the methods used to solve the failures corresponding to the first data. The output first data contains various punctuation marks and meaningless words, such as "de", "di", and ",", etc. Therefore, these words and punctuation marks need to be removed first. At the same time, when obtaining data, existing equipment or methods can be used, such as measuring the current in the circuit of the equipment through a sensor. As shown in the following table, it is part of the failures collected through the first data and the methods used to handle the corresponding failures, that is, the second data. The knowledge base interface formed later is as Figure 2 shown. For example, according to the following table, when the digital mixer has a failure log of "device network has been disconnected", that is, the first data, at this time, the second data on the method used to solve the failure corresponding to the first data is the content corresponding to the first data in the processing plan, that is, "restart the device, network jitter".

[0072]

[0073] S2. Calculate the total importance of each word segment in the first data, extract several word segments with the highest total importance as feature words, and establish a knowledge base for the corresponding relationship between the first data and the second data according to the feature words; each extracted word segment may also contain words that do not have a distinguishing meaning, such as words like "abnormal", etc. Each first data may contain, or most first data contain this word segment. Therefore, in order to save the calculation amount, this part of the words can be removed first, and then select the word segments with higher importance and those that can reflect the particularity of the first data from the remaining word segments. Therefore, in step S2, it specifically includes the following steps:

[0074] S21. Set a custom word library according to the first data, and perform word segment extraction on the first data respectively to obtain several word segment libraries, and set part-of-speech tags for the extracted word segments; the expression of the word segment library is: C = [c1, c2,..., c i], where C represents the word segmentation library, c n It represents the i-th word in the word library. The word extraction method can be performed through a word segmentation tool. At the same time, the word library for word segmentation extraction can be a professional dictionary or a modern Chinese dictionary, etc. However, based on the English appearing in the first data, it is added to the dictionary to form a custom word library to facilitate the subsequent extraction of words in the first data.

[0075] S22, set a first window, traverse each word in each word library in turn, and set connecting edges between the words that appear in the same first window at the same time to obtain a connecting edge set; the expression of the connecting edge set is: L = [l 1，2 , l 1，3 , ..., l m-1，m ], and then obtain the word segmentation graph according to the connection edge set and the word segmentation library, G = (C, L), where G represents the word segmentation graph, L represents the connection edge set, C represents the word segmentation library, and l i-1，i Indicates that there is a connecting edge between the i-th word and the i-1-th word in the word library;

[0076] S23. According to the set of connection edges, several segmented words with the largest number of connection edges are obtained to obtain a set of key words. The larger the number of connection edges, the greater the importance of the segmented words. Therefore, in order to reduce the amount of subsequent calculations, the segmented words are further screened. For example, in the process of outputting the first data, the first data may contain device information, such as the device number. The relationship between the number and the cause of the fault is relatively weak, so the segmented words of this type may be eliminated.

[0077] S24, calculate the frequency importance and overall importance of each word in the key word set, and calculate the first importance according to the product of the two; the first importance and the number of times each word in the key word set appears in the word library to be divided, establish the relationship between the number and the first importance, as the basis for selecting subsequent feature words, in step S24, specifically including the following steps:

[0078] S241, sequentially obtain the number of occurrences of any word in the key word set in the corresponding word library and the total number of occurrences of all words in the word library, and calculate the ratio of the two to obtain the frequency importance of the word; the calculation formula of the frequency importance is:

[0079]

[0080] In the above formula, Z p Indicates the frequency importance of the word. Indicates that the i-th word is in the corresponding word library m i The number of times it appears in Indicates the number of times the jth word appears in the word library. Represents the word segmentation library m iThe total number of occurrences of all participles in ;

[0081] S242, according to the total number of word segments and whether each word segment contains any word, calculate the distinction importance of the word segment, and calculate the distinction importance of each word segment in the key word set in turn; the calculation formula of the distinction importance is:

[0082]

[0083] In the above formula, Z q Indicates the importance of each word in the key word set, n m Indicates the total number of word libraries, count(c j ∈m) means that the statistics contain word c j The total number of word libraries;

[0084] S243, calculating the first importance according to the product of the word frequency importance and the distinction importance; the calculation formula of the first importance is:

[0085] Z1=Z p ×Z q

[0086] In the above formula, Z1 represents the first importance, Z p Indicates the frequency importance of the word, Z q Indicates the importance of distinction;

[0087] S25, according to the occurrence probability of any word in the key word set in the corresponding word library and the total number of occurrences of all words in the word library, the confusion degree of the word is calculated, and the second importance is calculated according to the part of speech of the word and the confusion degree; in a fault log, the parts of speech of different words, such as verbs and nouns, are of the highest importance, so as to improve the importance of words with important parts of speech such as verbs and nouns, and thus realize the calculation of the second importance. In step S25, the following steps are specifically included:

[0088] S251. Calculate the chaos degree of any word in the key word set according to the probability of occurrence of the word in the corresponding word library and the total number of occurrences of all words in the word library. The calculation formula of the chaos degree is:

[0089]

[0090] In the above formula, H(c i ) indicates the focus word concentration participle c i The degree of disorder, n m Indicates the total number of occurrences of all words in all word libraries. Indicates the focus word concentration participle c i The total number of occurrences in all word segments;

[0091] S252, calculating the second importance of each participle in the key word set according to the confusion degree and the part of speech of each participle in the key word set; the calculation formula of the second importance is:

[0092] Z2=γ×H(c i )

[0093] in,

[0094]

[0095] In the above formula, Z2 is the second importance, γ is the part-of-speech coefficient, γ1, γ2 and γ3 are the first part-of-speech coefficient, the second part-of-speech coefficient and the third part-of-speech coefficient respectively. After experiments, when γ1, γ2 and γ3 are set to 0.5, 0.4 and 0.1 respectively, the subsequent output second data is relatively more accurate, and the reliability of the output result is relatively high.

[0096] S26, calculate the third importance of any word according to its position in the key word set; the arrangement order of the words in the key word set corresponds to the arrangement order of the words in the corresponding first data, and the order of appearance of different words may mean that the corresponding first data is different, and the same order means that the probability of the information of the two fault logs being the same is greater, so it is necessary to calculate the third importance to reflect it, in step S26, specifically including the following steps:

[0097] S261, constructing a connection matrix of the key word set; the expression of the connection matrix is:

[0098]

[0099] In the above formula, M Q Represents the connection matrix of the key word set, G k,k-1 Indicates whether there is a connecting edge between the kth word and the k-1th word in the key word set. If there is a connecting edge, the result is 1, otherwise it is 0;

[0100] S262, calculating the third importance of any word in the key word set according to the connection matrix;

[0101]

[0102] In the above formula, Z3 represents the third importance of any word in the key word set. and Respectively represent the positions of the i-th and j-th words in the key word set, and the results are i and j respectively, where This is usually the position of the word with the largest number of connecting edges.

[0103] S27, calculating the total importance of any word in the key word set according to the first importance, the second importance and the third importance; the calculation formula of the total importance is:

[0104] Z 0 =A×Z1+B×Z2+C×Z3

[0105] In the above formula, Z 0 represents the total importance, A, B and C represent the first weight coefficient, the second weight coefficient and the third weight coefficient respectively, where A+B+C=1.

[0106] S28. Obtain several word segments with the largest total importance values ​​in each word segment library in turn, mark them as feature words, and correspond them to the second data to build a knowledge base. Therefore, each fault log, that is, the first data, corresponds to the second data through the feature words, and the feature words serve as the basis for subsequent selection of the second data.

[0107] S3, obtaining the third data including the current fault log, and determining whether there is the first data identical to the third data; the contents mentioned in step S1 to step S2 are the construction of the knowledge base, which is the preparatory work. At this time, the third data is obtained through the contents mentioned in step S3. Once the third data is identical to a previous first data, the corresponding second data is output, and the corresponding processing method can be obtained;

[0108] If yes, then output the second data corresponding to the first data according to the knowledge base and proceed to step S4;

[0109] If not, proceed to step S5;

[0110] S4, judging whether the second data has completed processing the third data, that is, whether the obtained second data can solve the fault recorded in the third data, because even if the third data has the same content as the first data, the causes of the fault may be different. In this case, it is necessary to further use the reference method provided in step S5 for reference by the operation and maintenance personnel, so as to assist the operation and maintenance personnel to quickly solve the fault;

[0111] If so, then end;

[0112] If not, proceed to step S5;

[0113] S5, obtaining feature words of the third data, converting the feature words in the first data and the third data into vectors, calculating the fit between the third data and each first data, and outputting the second data corresponding to a number of first data with the greatest fit; in order to further clarify the calculation method of the fit, step S5 specifically includes the following steps:

[0114] S51, obtaining feature words of the third data, and converting the feature words in the first data and the third data into vectors, for example, using a word embedding model Word2vec, which can train word vectors in conjunction with context information, and converting the feature words in the first data and the third data into vectors;

[0115] S52. Calculate the fit between the first data and the third data according to the vector of the feature word; the calculation formula of the fit is:

[0116]

[0117] In the above formula, TH represents the fit between the first data and the third data, I ik and I jk Represents the vector of the kth feature word in the first data and the third data respectively. There are P feature words in total. T i is a times weight coefficient, which sets the times of output of the second data corresponding to the first data, and the times weight coefficient is positively correlated with the times of output;

[0118] S53, extracting a plurality of first data with the greatest fit, and outputting the corresponding second data.

[0119] When the present invention is implemented, first, first data containing several historical fault logs is used to extract the segmented words in the first data, and then, by gradually calculating the first importance, the second importance and the third importance, several segmented words are screened out from various aspects as dimensions to form several feature words relative to the first data. Subsequently, when the third data is obtained, if there is first data that is the same as the third data, the corresponding second data is directly output as a processing method. Once the second data cannot solve the fault recorded in the third data, at this time, the final fit calculation is used to provide the operation and maintenance personnel with the second data for reference, so as to assist the operation and maintenance personnel in quickly solving the fault and ensure the normal progress of the video conference.

[0120] The above implementation methods have been described in detail. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A video conferencing equipment fault diagnosis method based on a knowledge base, characterized in that: The diagnostic method comprises the following steps: S1, obtaining a plurality of first data including historical fault logs of video conferencing equipment, and second data of a method used to solve the fault corresponding to the first data; S2, calculating the total importance of each word in the first data, extracting several words with the highest total importance as feature words, and establishing a knowledge base of the corresponding relationship between the first data and the second data based on the feature words; S3, obtaining third data including the current fault log, and determining whether there is first data identical to the third data; If yes, then output the second data corresponding to the first data according to the knowledge base and proceed to step S4; If not, proceed to step S5; S4, determining whether the second data has completed processing the third data; If so, then end; If not, proceed to step S5; S5. Obtain feature words of the third data, convert the feature words in the first data and the third data into vectors, calculate the degree of fit between the third data and each first data, and output the second data corresponding to a number of first data with the greatest degree of fit.

2. The diagnostic method according to claim 1, characterized in that In step S2, the following steps are specifically included: S21, setting a custom word library according to the first data, and performing word segmentation extraction on the first data respectively to obtain a plurality of word segmentation libraries, and setting part-of-speech tags for the extracted word segmentations; S22, setting a first window, traversing each word in each word library in turn, and setting connecting edges between the words that appear in the same first window at the same time to obtain a connecting edge set; S23, obtaining a number of word segments with the largest number of connection edges according to the connection edge set to obtain a key word set; S24, calculating the frequency importance and overall importance of each word in the key word set, and calculating the first importance according to the product of the two; S25, calculating the confusion degree of any word in the key word set according to the probability of occurrence of the word in the corresponding word library and the total number of occurrences of all words in the word library, and calculating the second importance according to the part of speech of the word and the confusion degree; S26, calculating the third importance of any participle according to its position in the key word set; S27, calculating the total importance of any word in the key word set according to the first importance, the second importance, and the third importance; S28. Obtain several word segments with the largest total importance values ​​in each word segment library in turn, mark them as feature words, and correspond them to the second data to construct a knowledge base.

3. The diagnostic method according to claim 2, characterized in that In step S24, the following steps are specifically included: S241, sequentially obtain the number of occurrences of any word in the key word set in the corresponding word library and the total number of occurrences of all words in the word library, and calculate the ratio of the two to obtain the frequency importance of the word; the calculation formula of the frequency importance is: In the above formula, Z p Indicates the frequency importance of the word. Indicates that the i-th word is in the corresponding word library m i The number of times it appears in Indicates the number of times the jth word appears in the word library. Represents the word segmentation library m i The total number of occurrences of all participles in ; S242, calculating the distinction importance of a segmented word according to the total number of segmented word libraries and whether each segmented word library contains any segmented word, and calculating the distinction importance of each segmented word in the key word set in turn; S243, calculating the first importance according to the product of the word frequency importance and the distinction importance; the calculation formula of the first importance is: Z1=Z p ×Z q In the above formula, Z1 represents the first importance, Z p Indicates the frequency importance of the word, Z q Indicates the importance of distinction.

4. The diagnostic method according to claim 2, characterized in that: In step S25, the following steps are specifically included: S251. Calculate the chaos degree of any word in the key word set according to the probability of occurrence of the word in the corresponding word library and the total number of occurrences of all words in the word library. The calculation formula of the chaos degree is: In the above formula, H(c i ) indicates the focus word concentration participle c i The degree of disorder, n m Indicates the total number of occurrences of all words in all word libraries. Indicates the focus word concentration participle c i The total number of occurrences in all word segments; S252: Calculate the second importance of each word in the key word set according to the confusion degree and the part of speech of each word in the key word set.

5. The diagnostic method according to claim 2, characterized in that: In step S26, the following steps are specifically included: S261, constructing a connection matrix of the key word set; the expression of the connection matrix is: In the above formula, M Q Represents the connection matrix of the key word set, G k,k-1 Indicates whether there is a connecting edge between the kth word and the k-1th word in the key word set. If there is a connecting edge, the result is 1, otherwise it is 0; S262, calculating the third importance of any word in the key word set according to the connection matrix; In the above formula, Z3 represents the third importance of any word in the key word set. and Respectively represent the positions of the i-th and j-th words in the key word set, and the results are i and j respectively, where is the position of the word with the largest number of connecting edges.

6. The diagnostic method according to claim 1, characterized in that: In step S5, the following steps are specifically included: S51, obtaining feature words of the third data, and converting the feature words in the first data and the third data into vectors; S52, calculating the degree of fit between the first data and the third data according to the vector of the feature word; S53, extracting a plurality of first data with the greatest fit, and outputting the corresponding second data.

7. The diagnostic method according to claim 3, characterized in that: In step S242, the calculation formula of the difference importance is: In the above formula, Z q Indicates the importance of each word in the key word set, n m Indicates the total number of word segmentation libraries, count(c j ∈m) means that the statistics contain word c j The total number of word libraries.

8. The diagnostic method according to claim 4, characterized in that In step S252, the calculation formula of the second importance is: Z2=γ×H(c i ) in, In the above formula, Z2 is the second importance, γ is the part-of-speech coefficient, γ1, γ2 and γ3 are the first part-of-speech coefficient, the second part-of-speech coefficient and the third part-of-speech coefficient respectively.

9. The diagnostic method according to claim 8, characterized in that: In step S252, γ1, γ2 and γ3 are set to 0.5, 0.4 and 0.1 respectively.

10. The diagnostic method according to claim 6, characterized in that: In step S52, the calculation formula of the fit degree is: In the above formula, TH represents the fit between the first data and the third data, I ik and I jk Represents the vector of the kth feature word in the first data and the third data respectively. There are P feature words in total. T i is a frequency weight coefficient, and sets the output frequency of the second data corresponding to the first data. The frequency weight coefficient is positively correlated with the output frequency.

Images