Method and system for establishing a data correlation network for the automotive industry

Abstract

Embodiments of the present application provide a method and system for establishing an automobile industry data correlation network. The method comprises: unifying each field name of each item of data in an automobile industry database to obtain a field vocabulary set; comparing the data field names with the field vocabulary set to obtain a field name vocabulary table; calculating a field name similarity matrix through the field name vocabulary table, calculating an application scenario similarity matrix of each item of data in the automobile industry database according to an application scenario, and performing weighted summation of the name similarity matrix and the application scenario similarity matrix to obtain a data correlation degree matrix; determining a data correlation relationship matrix according to the data correlation degree matrix, and forming an automobile industry data correlation network from the data correlation relationship matrix. Embodiments of the present application determine the correlation of data from two dimensions of data content and data application scenario, and establish an automobile data correlation network.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of data processing, and particularly relates to a method and system for establishing an automobile industry data correlation network. BACKGROUND

[0002] At present, the establishment of data correlation is mainly based on correlation rule related algorithms, the most classic of which are Apriori algorithm and FP-Growth algorithm.

[0003] 1. The main idea of the Apriori algorithm is as follows: (1) find frequent item sets, use the property that non-empty subsets of frequent item sets are frequent item sets, find the candidate set of frequent item sets first, count in the candidate set, and the frequent set greater than the support is the frequent set; (2) generate association rules from the frequent item set. The reference is Agrawal R. Srikant R. Fast algorithms for mining association rules. In Proceedings of the 20th International Conference on Very Large Databases [M]. Santiago, Chile, 1994: 487-499.

[0004] 2. The main idea of the FP-Growth algorithm is as follows: adopt the divide-and-conquer strategy, compress the frequent set in the database into a frequent pattern tree (FP-Tree) after the first scan, while still retaining the associated information therein, then differentiate the FP-Tree into some conditional databases, each of which is associated with a frequent set of length 1, and then mine these conditional databases respectively. The reference is HAN J W, KAMBER M, PEI J. Data mining: concepts and techniques [M]. Fan Ming, Meng Xiaofeng translation. 3rd edition. Beijing: Mechanical Industry Press, 2017: 158-169.

[0005] With the development of the new four modernizations of automobiles, the automobile industry data is growing exponentially, and the establishment of the automobile industry data correlation network can effectively solve the problem of data islands and is the basis for the value mining of automobile big data. At present, there is no method for establishing an automobile industry data correlation network, and the automobile industry database has problems such as a large number of data fields, non-standard names, and wide data ranges. The traditional data correlation is for the establishment of data correlation in the same range, and the direct use of the traditional data correlation for the establishment of the automobile industry data correlation network has poor effect. SUMMARY

[0006] The method and system for establishing a data association network in the automotive industry proposed in this invention unifies the field names in each data item in the automotive industry database, determines the data association from two dimensions: data content and data application scenario, and establishes an automotive data association network.

[0007] In a first aspect, embodiments of the present invention propose a method for establishing a data association network in the automotive industry, comprising:

[0008] Unified automotive industry database D = {D1, D2, ... D} N The names of each field in each data item are used to obtain the field vocabulary set {C'1,C'2,…,C'}. n}; Extract the field names of various data from the automotive industry database, compare them with the field names in the field vocabulary set, and compile a field name vocabulary L. N×n ;

[0009] Through the field name vocabulary L N×n Calculate the similarity matrix E1 of the data field names in the automotive industry database. N×N Based on the application scenarios of various data in the automotive industry database, calculate the similarity matrix E2 of the application scenarios of various data in the automotive industry database. N×N The name similarity matrix E1 N×N Similarity matrix with application scenarios E2 N×N Weighted summation yields the data correlation matrix E of the automotive industry database. N×N ;

[0010] According to the data correlation matrix E N×N Determine the data relationship matrix E' of the various data in the automotive industry database. N×N The data association matrix E' N×N To form a data network linking the automotive industry.

[0011] Preferably, the field vocabulary set {C'1,C'2,…,C'} is obtained. n The steps include:

[0012] Extracting the automotive industry database D = {D1, D2, ... D} N The names of all fields in all data in} form the initial field vocabulary set {C1, C2, ..., C}. m};

[0013] The field name character set of the initial field vocabulary is obtained by splitting each field name in the initial field vocabulary by character.

[0014] If the proportion of the number of identical characters in the character sets of two field names to the total number of characters in their respective field names is greater than th1, replace the field name with the one containing fewer characters with the one containing more characters. If the number of identical characters in the two field names is the same, unify the two field names into one of the two field names; otherwise, do nothing.

[0015] After removing duplicate field names, the merged field name set {C1, C2, ..., C} is obtained. m-1};

[0016] Repeat the above steps until the proportion of the number of identical characters in the names of any two field names in the merged field vocabulary is less than or equal to th1, resulting in the field vocabulary {C'1, C'2, ..., C'...}. n}

[0017] Preferably, the threshold th1 is used to determine the similarity between two field names, and its value ranges from 0.5 to 0.75.

[0018] Preferably, the automotive industry database field name vocabulary list L is obtained. N×n The steps include:

[0019] Extract all field names and field vocabulary {C'1, C'2, ..., C'} of the i-th data item in the automotive industry data. n The c-th field name in the data is compared. If the i-th data item in the automotive industry data contains a field name that matches the field name C' in the field vocabulary set... c ={C'1,C'2,…C' dc If the proportion of the number of identical characters in the field name of the i-th data item in the automotive industry data is greater than th1, then L i×c =1, otherwise L i×c =0;

[0020] Iterate through the vocabulary set {C'1, C'2, ..., C'} using the methods described above. n} This retrieves all field names from the} database, obtaining the i-th data item in the automotive industry database corresponding to the i-th row of the automotive industry database field name vocabulary list L. i Ultimately, the automotive industry database field name vocabulary L was obtained. N×n .

[0021] Preferably, calculate the field name similarity matrix E1. N×N The steps include:

[0022] Using Euclidean distance and field name vocabulary L N×n Calculate D in the automotive industry database k D jThe field name similarity E1 of the two sub-databases (k,j) That is,

[0023]

[0024] According to the above steps, the field name similarity matrix E1 of each item of data in the automobile industry database is calculated N×N .

[0025] Preferably, the step of calculating the application scenario similarity matrix E2 N×N includes:

[0026] Tokenizing the application scenario description statement of each item of data in the automobile industry database to obtain the word vector of the application scenario description language of each item of data;

[0027] Fusing and deduplicating the word vectors of all data to obtain a total set of word vectors of the automobile industry database;

[0028] According to the word vector and the total set of word vectors, a term frequency vector set F N×M is obtained according to a term frequency-inverse document frequency model;

[0029] The field name similarity E2 of the two items of data D k , D j in the automobile industry database is calculated by the Euclidean distance (k,j) , that is,

[0030]

[0031] The above steps are repeated to obtain the application scenario similarity matrix E2 of each item of data in the automobile industry database N×N .

[0032] Preferably, the calculation method of the data correlation degree matrix E N×N is:

[0033] E N×N = λ1E1 N×N + λ2E2 N×N , where λ1+ λ2=1 and λ1≠0 and λ2≠0.

[0034] Preferably, the calculation method of the data correlation relationship matrix E' N×N is: when E (e,f) >th2, the data item D e is correlated with the data item D f , E' (e,f) =1; otherwise, E' (e,f) =0; the above steps are repeated to obtain the data correlation relationship matrix E' N×N .

[0035] Preferably, the threshold value th2 is used to determine whether the two data are associated, and the threshold value th2 is in the range of 0.6-0.7.

[0036] In a second aspect, the embodiment of the present application provides a system for establishing an automobile industry data association network, comprising:

[0037] A data field standardization module, which unifies each field name of each data in an automobile industry database D={D1, D2, … D N} to obtain a field vocabulary set {C'1, C'2, …, C' n}; The field name of each data in the automobile industry database is extracted, compared with the field name in the field vocabulary set, and a field name vocabulary L N×n is obtained by statistics.

[0038] A data association degree matrix solving module, which calculates a field name similarity matrix E1 N×n of each data in the automobile industry database according to the field name vocabulary L N×N , calculates an application scene similarity matrix E2 N×N of each data in the automobile industry database according to the application scene of each data in the automobile industry database, and obtains a data association degree matrix E N×N of each data in the automobile industry database by weighted summation of the name similarity matrix E1 N×N and the application scene similarity matrix E2 N×N .

[0039] A data association network establishing module, which determines a data association relationship matrix E' N×N of each data in the automobile industry database according to the data association degree matrix E N×N , forms an automobile industry data association network according to the data association relationship matrix E' N×N .

[0040] Compared with the prior art, the embodiment of the present application has the following specific beneficial effects:

[0041] 1. Unifying each field name of each data in the automobile industry database provides technical support for management based on master data.

[0042] 2. Establishing an automobile data association network realizes association analysis based on data content and data application scene, explores new application scenes through association, provides more data options for business development, improves data utilization rate, and accelerates data value release. BRIEF DESCRIPTION OF DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the specific embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the specific embodiments or prior art description. Obviously, the drawings described below are some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor based on these drawings.

[0044] Figure 1 The flowchart for establishing the automobile industry association network of the embodiment of the present application. DETAILED DESCRIPTION

[0045] In order to make the objectives, technical solutions and advantages of the present application clearer, the technical solutions of the present application will be described clearly and completely below. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of the present application.

[0046] The embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0047] In combination with Figure 1 , the embodiment of the present application proposes a method for establishing an automobile industry association network, and the specific steps include:

[0048] Step S100, unifying each field name of each item of data in the automobile industry database D={D1, D2, …D N} to obtain a field vocabulary set {C'1, C'2, …, C' n}; extracting the field names of each item of data in the automobile industry database, comparing with the field names in the field vocabulary set {C'1, C'2, …, C' n}, and statistically obtaining a field name vocabulary L N×n .

[0049] Specifically, the step of unifying each field name of each item of data in the automobile industry database to obtain the field vocabulary set {C'1, C'2, …, C' n} includes:

[0050] Step S111, extracting all field names of all data in the automobile industry database D={D1, D2, …D N}, to form an initial field vocabulary set {C1, C2, …, C m}.

[0051] Wherein, D1, D2, …D N are data in the automobile industry database, N is the total number of data in the automobile industry database; C1, C2, …C mis the field name of the data in the automobile industry database, m is the total number of field names of all data in the automobile industry database.

[0052] Step S112, each field name in the initial field vocabulary set is segmented by character to obtain a field name character set of the initial field vocabulary set

[0053] wherein, is the field name character set C a contains characters, d a is the field name character set C a contains the number of characters.

[0054] Step S113, if the proportion of the number of characters contained in two field names to the number of characters contained in itself is greater than th1, replace the field name with more characters with the field name with less characters, if the number of characters contained in two field names is the same, then unify the two field names into one of the two field names; otherwise, do not process, that is:

[0055] If and d a <d b , then C b =C a .

[0056] If and d a =d b , then C b =C a or C a =C b .

[0057] wherein, the threshold value th1 is used to judge the similarity of two field names, and the value range is between 0.5 and 0.75.

[0058] Step S114, remove the repeated field names to obtain a merged field vocabulary set {C1, C2, …, C m-1}.

[0059] Step S115, repeat steps S113 and S114 until the proportion of the number of characters contained in any two field names in the merged field vocabulary set to the total number of characters contained in itself is less than or equal to th1, stop repeating, and obtain a field vocabulary set {C'1, C'2, …, C' n}, wherein n is the total number of field names of all data in the final automobile industry database.

[0060] Specifically, extract the field names of various data items from the automotive industry database, and match them with the field vocabulary {C'1,C'2,…,C'...}. n By comparing the field names in}, a vocabulary of field names for the automotive industry database, L, was obtained. N×n The steps include:

[0061] Step S121: Extract all field names and field vocabulary {C'1,C'2,…,C'} of the i-th data item in the automotive industry data. n The c-th field name in the data is compared. If the i-th data item in the automotive industry data contains a field name that matches the field name in the field vocabulary set... If the proportion of the number of identical characters in the name of the i-th data field in the automotive industry data is greater than th1, then L i×c =1, otherwise L i×c =0.

[0062] Step S122: Apply the method from step S121 to traverse the field vocabulary set {C'1,C'2,…,C'} n} This retrieves all field names from the} database, obtaining the i-th data item in the automotive industry database corresponding to the i-th row of the automotive industry database field name vocabulary list L. i .

[0063] Step S123: Repeat steps S121 and S122 to obtain the automotive industry database field name vocabulary L. N×n Where N is the total number of data items in the automotive industry database, and n is the total number of field names in all data in the final automotive industry database.

[0064] Step S200, through the field name vocabulary L N×n Calculate the similarity matrix E1 of the data field names in the automotive industry database. N×N Based on the application scenarios of various data in the automotive industry database, calculate the similarity matrix E2 of the application scenarios of various data in the automotive industry database. N×N The name similarity matrix E1 N×N Similarity matrix with application scenarios E2 N×N Weighted summation yields the data correlation matrix E of the automotive industry database. N×N .

[0065] Specifically, through the field name vocabulary L N×n Calculate the similarity matrix E1 of the data field names in the automotive industry database. N×N The steps include:

[0066] Step S211, using Euclidean distance and field name vocabulary L N×n Calculate D in the automotive industry databasek D j Field name similarity E1 between the two sub-databases (k,j) ,Right now:

[0067]

[0068] Step S212: Following the method described above, calculate the similarity matrix E1 of the data field names in the automotive industry database. N×N Where N is the total number of data items in the automotive industry database.

[0069] Specifically, based on the application scenarios of various data in the automotive industry database, a similarity matrix E2 for the application scenarios of various data in the automotive industry database is calculated. N×N The steps include:

[0070] Step S221: The application scenario description information S = {S1, S2, ... S...} from the automotive industry database is used to generate various data. N The jieba library (a Chinese word segmentation library for the Pathol language) is used for word segmentation, removing punctuation and stop words to obtain the word vector S1 = {s1, s2, ...} for each data application scenario description language. It is understood that word segmentation libraries are not limited to jieba.

[0071] Step S222: Process all application scenario description statements in the automotive industry database according to step S221, and merge and deduplicate the word vectors of all data to obtain the total set of word vectors in the automotive industry database V = {v1, v2, ... v...} M}, where M is the total number of words in the word vector set of the automotive industry database.

[0072] Step S223: Based on the word vectors of all data application scenario description languages ​​and the total set of word vectors in the automotive industry database, and following the TF-IDF (Term Frequency-Inverse Text Frequency) method, obtain the word frequency vector set F. N×M .

[0073] Step S224: Calculate D in the automotive industry database using Euclidean distance. k D j Similarity of field names between two data items (E2) (k,j) ,Right now:

[0074]

[0075] Repeat the above steps to obtain the application scenario similarity matrix E2 of various data in the automotive industry database. N×N .

[0076] Specifically, the name similarity matrix E1 N×NThe application scene similarity matrix E2 N×N The weighted sum is the data correlation degree matrix E of the data in the automobile industry database. N×N The calculation method is as follows:

[0077] E N×N = λ1E1 N×N + λ2E2 N×N , wherein λ1+ λ2=1 and λ1≠0 and λ2≠0.

[0078] In step S300, the data correlation degree matrix E N×N is used to determine the data correlation relationship matrix E' N×N of the data in the automobile industry database. N×N The data correlation network of the automobile industry is formed according to the data correlation relationship matrix E'

[0079] Specifically, when the correlation degree between two data in the automobile industry database is greater than a threshold th2, the two data are correlated, that is, when a data value in the data correlation relationship matrix E' N×N is greater than the threshold th2, the two data represented by the data value are correlated. The data correlation relationship matrix of the data in the automobile industry database is E' N×N , when E (e,f) >th2, the data item D e is correlated with the data item D f , E' (e,f) =1; otherwise, E' (e,f) =0; the above steps are repeated to obtain the data correlation relationship matrix E' N×N . The threshold th2 is used to determine whether the two data are correlated, and the threshold th2 is in the range of 0.6-0.7.

[0080] The Gephi tool is used to draw the data correlation network of the automobile industry. If the value of two data in the data correlation relationship matrix is 1, a line segment is drawn to connect the nodes representing the two data, otherwise, the nodes are not connected; the above steps are repeated to obtain the data correlation network.

[0081] Another embodiment of the application provides a system for establishing a data correlation network of the automobile industry, comprising:

[0082] A data field standardization module is used to unify the field names of each data in the automobile industry database D={D1, D2, …, D N} to obtain a field vocabulary set {C'1, C'2, …, C' n}. The field names of each data in the automobile industry database are extracted and compared with the field names in the field vocabulary set to obtain a field name vocabulary L.N×n ;

[0083] a data correlation degree matrix solving module, which solves a data correlation degree matrix E of the data in the automobile industry database through the field name vocabulary L N×n calculating a name similarity matrix E1 of each data field name in the automobile industry database N×N calculating a name similarity matrix E1 of each data field name in the automobile industry database N×N , according to the application scene similarity matrix E2 of each data in the automobile industry database N×N , according to the application scene similarity matrix E2 of each data in the automobile industry database N×N weighted summation, obtaining the data correlation degree matrix E of each data in the automobile industry database N×N ;

[0084] a data correlation network establishing module, which establishes a data correlation network of the automobile industry database according to the data correlation degree matrix E N×N determining a data correlation relationship matrix E' of each data in the automobile industry database N×N , according to the data correlation relationship matrix E' N×N forming a data correlation network of the automobile industry database.

[0085] The system for establishing a data correlation network of the automobile industry provided in the embodiment can execute the method for establishing a data correlation network of the automobile industry provided in any embodiment, and has the corresponding technical effects, which will not be described herein again.

[0086] It should be understood that the various forms of procedures shown above can be reordered, added or deleted. For example, each step described in the present application can be executed in parallel, sequentially or in a different order, as long as the desired results of the technical solutions disclosed in the present application can be achieved, which is not limited herein.

[0087] The above specific embodiments do not constitute a limitation on the protection scope of the present application. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent replacement and improvement within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A method of establishing a data correlation network for the automotive industry, characterized in that, Comprising: Unified automobile industry database D = {D1, D2, … D N} each data field name, get the field vocabulary set {C'1, C'2, …, C' n}; Extract the field name of each data in the automobile industry database, compare it with the field name in the field vocabulary set, and count to get the field name vocabulary L N×n ; D1, D2, … D N is the data in the automobile industry database, N is the total number of data in the automobile industry database; C'1, C'2, …, C' n is the field name, n is the total number of field names owned by all data in the final automobile industry database; Obtaining an automotive industry database field name glossary L N×n comprises: extracting all field names of the i-th item of the automobile industry data and comparing the c-th field name in the set of field names and field glossary {C'1, C'2,..., C'N}, if there is a field name in the i-th item of the automobile industry data that is identical to a field name in the set of field names and field glossary n , the L = 1; otherwise, the L i×c = 0. i×c = 1 if the proportion of the number of identical characters in the field name of the i-th item of the automobile industry data to the total number of characters in the field name of the i-th item of the automobile industry data is greater than th1; otherwise, the L i×c = 0. Apply the method in the above steps to traverse all field names in the field vocabulary set {C'1, C'2, …, C'N} to obtain the automobile industry database field name vocabulary table L n corresponding to the i-th item of automobile industry data i , and finally obtain the automobile industry database field name vocabulary table L N×n ; Through the field name vocabulary L N×n Calculate the similarity matrix E1 of each data field name in the automobile industry database N×N According to the application scene of each data in the automobile industry database, calculate the similarity matrix E2 of the application scene of each data in the automobile industry database N×N , the name similarity matrix E1 N×N And the application scene similarity matrix E2 N×N Weighted sum, get the data correlation matrix E of each data in the automobile industry database N×N ; computing a field name similarity matrix E1 N×N comprises the steps of: By the Euclidean distance and the field name vocabulary L N×n D k , D j The field name similarity of the two sub-databases E1 (k,j) , i.e.: According to the method of the above steps, the similarity matrix E1 of the data field names in the automobile industry database is calculated N×N ; According to the data correlation matrix E N×N The data correlation relationship matrix E of each data of the automobile industry database is determined N×N The data correlation relationship matrix E of each data of the automobile industry database is determined N×N The automobile industry data correlation network is formed.

2. The method of claim 1, wherein, Get the vocabulary set {C'1,C'2,…,C'} n The steps include: Extract all field names of all data in the database D = {D1, D2, … D N} of the automotive industry to form the initial field vocabulary set {C1, C2, …, C m}. segmenting each field name in the initial field vocabulary set {C1, C2, …, C m} by characters to obtain a field name character set of the initial field vocabulary set If the proportion of the same character number in the character set of two field names to the character number contained in the field name itself is greater than th1, replace the field name with more character number contained in the two field names with the field name with less character number contained, if the character number contained in the two field names is the same, unify the two field names into one of the two field names; otherwise, do not process; Remove the duplicate field names, and get the merged field vocabulary {C1, C2, …, C m-1} Repeat the above steps until the proportion of the same character number contained in any two field names in the merged field vocabulary set to the total number of characters contained in itself is less than or equal to th1 Stop repeating, and get the field vocabulary set {C'1, C'2, …, C' n}.

3. The method of claim 2, wherein, The threshold th1 is used to judge the similarity of the two field names, and the value range is between 0.5 and 0.

75.

4. The method of claim 1, wherein, computing an application scenario similarity matrix E2 N×N The steps of the method include: The word vectors of each data application scene description language in the automobile industry database are obtained by segmenting the data application scene description language. All data word vectors are fused and de-duplicated to obtain a total set of automobile industry database word vectors. According to the word vector and the total set of word vectors, a set of word frequency vectors F is obtained according to a term frequency-inverse document frequency model N×M ; The Euclidean distance is calculated between D k , D j Two data field name similarity E2 (k,j) , namely: Repeat the above steps to obtain the similarity matrix E2 of each data application scenario in the automobile industry database N×N .

5. The method of claim 4, wherein, Data correlation matrix E N×N The calculation method is as follows: E N×N = λ1E1 N×N + λ2E2 N×N where λ1+ λ2= 1 and λ1≠ 0 and λ2≠ 0.

6. The method of claim 1, wherein, Data association relationship matrix E' N×N The calculation method is as follows: when E (e,f) >th2, the data item D e is associated with the data item D f , E' (e,f) =1; otherwise, E' (e,f) =0; repeat the above steps to obtain the data association relationship matrix E' N×N .

7. The method of claim 6, wherein, The threshold th2 is used to determine whether two data are associated, and the threshold th2 value range is 0.6-0.

7.

8. A system for establishing a data correlation network in the automotive industry, characterized in that Comprising: A data field standardization module unifies each field name of each data in the automotive industry database D = {D1, D2, … D N} to obtain a field vocabulary set {C'1, C'2, …, C' n}; the field name of each data in the automotive industry database is extracted and compared with the field name in the field vocabulary set to obtain a field name vocabulary L N×n ; D1, D2, … D N are data in the automotive industry database, N is the total number of data in the automotive industry database; C'1, C'2, …, C' n are field names, and n is the total number of field names possessed by all data in the final automotive industry database; Obtaining an automotive industry database field name glossary L N×n comprises: extracting all field names of the i-th item of the automobile industry data and comparing the c-th field name in the set of field names and field glossary {C'1, C'2,..., C'N}, if there is a field name in the i-th item of the automobile industry data which is identical to a field name in the set of field names and field glossary n , then L = 1, otherwise L i×c = 0. i×c = 1 if the proportion of the number of identical characters in the field name of the i-th item of the automobile industry data to the total number of characters in the field name of the i-th item of the automobile industry data is greater than th1, then L i×c = 1, otherwise L i×c = 0. Apply the method in the above steps to traverse all field names in the field vocabulary set {C'1, C'2, …, C'N} to obtain the automobile industry database field name vocabulary table L n corresponding to the i-th item of automobile industry data i , and finally obtain the automobile industry database field name vocabulary table L N×n ; a data correlation degree matrix solving module, which is configured to solve a data correlation degree matrix E through the field name vocabulary L N×n calculating a field name similarity matrix E1 of each item of data in the automobile industry database N×N calculating a data application scenario similarity matrix E2 of each item of data in the automobile industry database according to data application scenarios of each item of data in the automobile industry database N×N weighting and summing the name similarity matrix E1 N×N and the application scenario similarity matrix E2 N×N to obtain a data correlation degree matrix E of each item of data in the automobile industry database N×N The step of calculating the field name similarity matrix E1 N×N includes: By the Euclidean distance and the field name vocabulary L N×n D k , D j The field name similarity of the two sub-databases E1 (k,j) , i.e.: According to the method of the above steps, the similarity matrix E1 of the data field names in the automobile industry database is calculated N×N ; a data association network establishing module, which establishes a data association network according to the data association degree matrix E N×N determining a data association relationship matrix E of each data in the automobile industry database N×N According to the data association relationship matrix E N×N forming a data association network of the automobile industry.

Images