Large model information mining method and system

Abstract

The invention relates to the field of big data, and particularly provides a big model information mining method and system, and the method comprises the following steps: S1, determining a dialogue information relationship based on a knowledge graph, carrying out the database construction and graph definition of dialogue information in the capital construction field, and selecting a knowledge graph algorithm to carry out the extension association of the structure of data information; and S2, mining data information by parallel similarity relationships under a big data technology, and on the basis of determining an isomorphic relationship between dialogue information, associating the similarity relationships by adopting the big data technology so as to provide support for subsequent data mining. Compared with the prior art, the big data technology can be adopted, the data obtaining efficiency can be optimized, the data mining, collecting and analyzing quality can be improved, and compared with a traditional data mining algorithm, higher mining precision is achieved.

Description

Technical Field

[0001] The present invention relates to the field of big data, and specifically provides a large model information mining method and system. Background Art

[0002] Traditional macroeconomic data collection methods have many limitations and challenges. First, macroeconomic data are usually released by relevant departments, and their update frequency is limited by the administrative mechanisms and data processing procedures of relevant departments, resulting in a high time lag in the data. Second, due to the complexity and diversity of economic activities, traditional data collection methods often cannot cover comprehensive economic information, and some data may be distorted due to sampling bias or insufficient samples. In addition, traditional data collection methods often require a lot of manpower and material resources, and are easily interfered by human factors, making it difficult to ensure data quality.

[0003] With the rapid development of information technology, the application of big data technology in the field of macroeconomic data has become a trend. With its massive, diverse, real-time and efficient characteristics, big data technology provides new ideas and methods for solving the problems of traditional data collection methods. Through big data mining technology based on knowledge graphs, real-time collection and analysis of economic data in various fields can be achieved, thereby improving the comprehensiveness and accuracy of data, shortening the data update cycle, and providing more accurate and timely support for economic decision-making.

[0004] Therefore, how to combine big data technology with macroeconomic data collection, analysis and prediction has become a problem in the current economic field. Summary of the invention

[0005] The present invention aims at the above-mentioned deficiencies of the prior art and provides a large model information mining method with strong practicality.

[0006] A further technical task of the present invention is to provide a large model information mining system that is reasonably designed, safe and applicable.

[0007] The technical solution adopted by the present invention to solve its technical problem is:

[0008] A large model information mining method has the following steps:

[0009] S1. Determine the relationship between conversation information based on the knowledge graph, build a database of conversation information in the infrastructure field and define the graph, and select the knowledge graph algorithm to expand and associate the structure of data information;

[0010] S2. Parallel similarity relationship mining data information under big data technology. On the basis of determining the isomorphic relationship between dialogue information, big data technology is used to relate similar relationships to provide support for subsequent data mining.

[0011] Furthermore, in step S1, only data that meets the support has the connected property. The support of each group of information in the data flow graph is obtained according to the multi-label model in the knowledge graph, and two relationship models are set respectively, represented as q and W.

[0012] Furthermore, in step S2, the two relationship modes generate three extension situations, as follows:

[0013] (1) The q relationship needs to be obtained by expanding the w relationship to the right;

[0014] (2) The q relation is obtained by backward extension of the w relation;

[0015] (3) The q relationship is obtained by extending the relationship forward;

[0016] According to the above-mentioned extended pattern, assuming that the data set with q relationship is e, and the data set with w relationship is e, inequalities are established for different situations. In the process of determining the relational pattern of dialogue information, it is necessary to obtain the intrinsic instance support M(y) of the graph pattern.

[0017] Further, in step S2, the acquired data information is constructed into a specific relationship set, according to the two sets of sets e established w , e q Mining is performed to verify whether the similarity merging of data is achieved. However, there are two probability situations in the similarity merging of data sets, namely, intersection probability and non-intersection probability. Corresponding to the whole life cycle of data mining, the combination situation is calculated in the global field:

[0018]

[0019] Among them, a(e w ), a(e q ) are two sets of sets e w , e q The global probability of

[0020] The parallel similarity mining process is described as follows:

[0021]

[0022] Among them, sim() is the similarity function, f is the attribute value of the instance data, a f is the probability of occurrence of f, s f is the weight of the data set that has attribute f in common, d f is the correction parameter;

[0023] Through the above methods, knowledge graph and big data technology are combined to complete data mining and analysis.

[0024] A large model information mining system, firstly, determines the relationship of dialogue information based on the knowledge graph, builds a database of dialogue information in the infrastructure field and defines the graph, and selects the knowledge graph algorithm to expand and associate the structure of data information;

[0025] Then, the data information is mined by parallel similar relationships under big data technology. On the basis of determining the isomorphic relationship between the dialogue information, big data technology is used to relate the similar relationships to provide support for subsequent data mining.

[0026] Furthermore, only data that meets the support requirements have the property of being connected. The support of each group of information in the data flow graph is obtained according to the multi-label model in the knowledge graph, and two relationship models are set respectively, represented as q and W.

[0027] Furthermore, the two relationship modes described above produce three expansion situations, as follows:

[0028] (1) The q relationship needs to be obtained by expanding the w relationship to the right;

[0029] (2) The q relation is obtained by backward extension of the w relation;

[0030] (3) The q relationship is obtained by extending the relationship forward;

[0031] According to the above-mentioned extended pattern, assuming that the data set with q relationship is e, and the data set with w relationship is e, inequalities are established for different situations. In the process of determining the relational pattern of dialogue information, it is necessary to obtain the intrinsic instance support M(y) of the graph pattern.

[0032] Furthermore, the acquired data information is constructed into a specific relationship set, according to the two sets of sets established w , e q Mining is performed to verify whether the similarity merging of data is achieved. However, there are two probability situations in the similarity merging of data sets, namely, intersection probability and non-intersection probability. Corresponding to the whole life cycle of data mining, the combination situation is calculated in the global field:

[0033]

[0034] Among them, a(e w ), a(e q ) are two sets of sets e w , e q The global probability of

[0035] The parallel similarity mining process is described as follows:

[0036]

[0037] Among them, sim() is the similarity function, f is the attribute value of the instance data, a f is the probability of occurrence of f, s f is the weight of the data set that has attribute f in common, d f is the correction parameter;

[0038] Through the above methods, knowledge graph and big data technology are combined to complete data mining and analysis.

[0039] Compared with the prior art, the large model information mining method and system of the present invention has the following outstanding beneficial effects:

[0040] The present invention adopts big data technology, can optimize the efficiency of data acquisition, improve the quality of data mining, collection and analysis, and has higher mining accuracy than traditional data mining algorithms. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0042] Attached Figure 1 It is a flowchart of a large model information mining method. DETAILED DESCRIPTION

[0043] In order to enable those skilled in the art to better understand the solution of the present invention, the present invention is further described in detail below in conjunction with specific implementation methods. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0044] A best embodiment is given below:

[0045] like Figure 1 As shown, a large model information mining method in this embodiment has the following steps:

[0046] S1. Determine the relationship between conversation information based on the knowledge graph, build a database of conversation information in the infrastructure field and define the graph, and select the knowledge graph algorithm to expand and associate the structure of data information.

[0047] Only data that meets the support degree has the property of connection. The support degree of each group of information in the data flow graph is obtained according to the multi-label mode in the knowledge graph. Two relationship modes are set respectively, represented as q and W. The above two relationship modes will produce three expansion situations, as follows:

[0048] (1) The q relationship needs to be obtained by expanding the w relationship to the right;

[0049] (2) The q relation is obtained by backward extension of the w relation;

[0050] (3) The q relationship is obtained by extending the relationship forward;

[0051] According to the above-mentioned extended pattern, assuming that the data set with q relationship is e, and the data set with w relationship is e, inequalities are established for different situations. In the process of determining the relational pattern of dialogue information, it is necessary to obtain the intrinsic instance support M(y) of the graph pattern.

[0052] S2. Parallel similarity relationship mining data information under big data technology. On the basis of determining the isomorphic relationship between dialogue information, big data technology is used to relate similar relationships to provide support for subsequent data mining.

[0053] The acquired data information is constructed into a specific relationship set, still according to the two sets of sets established above w , e q Mining is performed to verify whether it can realize the similarity merging of data. However, there are two probability situations in the similarity merging of data sets, namely, the intersection probability and the non-intersection probability. Corresponding to the whole life cycle of data mining, the combination situation is calculated in the global field:

[0054]

[0055] Among them, a(e w ), a(e q ) are two sets of sets e w , e q The global probability

[0056] The parallel similarity mining process is described as follows:

[0057]

[0058] Among them, sim() is the similarity function, f is the attribute value of the instance data, a f is the probability of occurrence of f, s f is the weight of the data set that has attribute f in common, d f To correct the parameters.

[0059] Through the above methods, knowledge graph and big data technology are combined to complete data mining and analysis.

[0060] Based on the above method, a large model information mining system in this embodiment first determines the relationship of dialogue information based on the knowledge graph, constructs a database of dialogue information in the infrastructure field and defines the graph, and selects the knowledge graph algorithm to expand and associate the structure of data information;

[0061] Then, the data information is mined by parallel similar relationships under big data technology. On the basis of determining the isomorphic relationship between the dialogue information, big data technology is used to relate the similar relationships to provide support for subsequent data mining.

[0062] Among them, only data that meets the support has the connected property. According to the multi-label model in the knowledge graph, the support of each group of information in the data flow graph is obtained, and two relationship models are set respectively, represented as q and W.

[0063] The two relationship modes described above produce three expansion situations, as follows:

[0064] (1) The q relationship needs to be obtained by expanding the w relationship to the right;

[0065] (2) The q relation is obtained by backward extension of the w relation;

[0066] (3) The q relationship is obtained by extending the relationship forward;

[0067] According to the above-mentioned extended pattern, assuming that the data set with q relationship is e, and the data set with w relationship is e, inequalities are established for different situations. In the process of determining the relational pattern of dialogue information, it is necessary to obtain the intrinsic instance support M(y) of the graph pattern.

[0068] The acquired data information is constructed into a specific relationship set, according to the two sets of sets established w , e q Mining is performed to verify whether the similarity merging of data is achieved. However, there are two probability situations in the similarity merging of data sets, namely, intersection probability and non-intersection probability. Corresponding to the whole life cycle of data mining, the combination situation is calculated in the global field:

[0069]

[0070] Among them, a(e w ), a(e q ) are two sets of sets e w , e q The global probability of

[0071] The parallel similarity mining process is described as follows:

[0072]

[0073] Among them, sim() is the similarity function, f is the attribute value of the instance data, a f is the probability of occurrence of f, s f is the weight of the data set that has attribute f in common, d f is the correction parameter;

[0074] Through the above methods, knowledge graph and big data technology are combined to complete data mining and analysis.

[0075] The above-mentioned specific implementations are only specific cases of the present invention. The patent protection scope of the present invention includes but is not limited to the above-mentioned specific implementations. Any technical solutions that conform to the above-mentioned specific implementations of the present invention and any appropriate changes or substitutions made by ordinary technicians in the relevant technical field shall fall within the patent protection scope of the present invention.

[0076] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A large model information mining method, characterized in that: The steps are as follows: S1. Determine the relationship between conversation information based on the knowledge graph, build a database of conversation information in the infrastructure field and define the graph, and select the knowledge graph algorithm to expand and associate the structure of data information; S2. Parallel similarity relationship mining data information under big data technology. On the basis of determining the isomorphic relationship between dialogue information, big data technology is used to relate similar relationships to provide support for subsequent data mining.

2. A large model information mining method according to claim 1, characterized in that: In step S1, only data that meets the support requirements have the property of being connected. The support of each group of information in the data flow graph is obtained according to the multi-label model in the knowledge graph, and two relationship models are set respectively, represented as q and W.

3. A large model information mining method according to claim 2, characterized in that: In step S2, the two relationship modes generate three expansion situations, as follows: (1) The q relationship needs to be obtained by expanding the w relationship to the right; (2) The q relation is obtained by backward extension of the w relation; (3) The q relationship is obtained by extending the relationship forward; According to the above-mentioned extended pattern, assuming that the data set with q relationship is e, and the data set with w relationship is e, inequalities are established for different situations. In the process of determining the relational pattern of dialogue information, it is necessary to obtain the intrinsic instance support M(y) of the graph pattern.

4. A large model information mining method according to claim 3, characterized in that: In step S2, the acquired data information is constructed into a specific relationship set, according to the two sets of sets e established w , e q Mining is performed to verify whether the similarity merging of data is achieved. However, there are two probability situations in the similarity merging of data sets, namely, intersection probability and non-intersection probability. Corresponding to the whole life cycle of data mining, the combination situation is calculated in the global field: Among them, a(e w ), a(e q ) are two sets of sets e w , e q The global probability of The parallel similarity mining process is described as follows: Among them, sim() is the similarity function, f is the attribute value of the instance data, a f is the probability of occurrence of f, s f is the weight of the data set that has attribute f in common, d f is the correction parameter; Through the above methods, knowledge graph and big data technology are combined to complete data mining and analysis.

5. A large model information mining system, characterized in that: First, we determine the relationship between conversation information based on the knowledge graph, build a database of conversation information in the infrastructure field, and define the graph. We then select the knowledge graph algorithm to expand and associate the structure of the data information. Then, the data information is mined by parallel similar relationships under big data technology. On the basis of determining the isomorphic relationship between the dialogue information, big data technology is used to relate the similar relationships to provide support for subsequent data mining.

6. A large model information mining system according to claim 5, characterized in that: Only data that meets the support degree has the property of connection. The support degree of each group of information in the data flow graph is obtained according to the multi-label mode in the knowledge graph, and two relationship modes are set respectively, represented as q and W.

7. A large model information mining system according to claim 6, characterized in that: The two relationship modes described above produce three expansion situations, as follows: (1) The q relationship needs to be obtained by expanding the w relationship to the right; (2) The q relation is obtained by backward extension of the w relation; (3) The q relationship is obtained by extending the relationship forward; According to the above-mentioned extended pattern, assuming that the data set with q relationship is e, and the data set with w relationship is e, inequalities are established for different situations. In the process of determining the relational pattern of dialogue information, it is necessary to obtain the intrinsic instance support M(y) of the graph pattern.

8. A large model information mining system according to claim 7, characterized in that: The acquired data information is constructed into a specific relationship set, according to the two sets of sets established w , e q Mining is performed to verify whether the similarity merging of data is achieved. However, there are two probability situations in the similarity merging of data sets, namely, intersection probability and non-intersection probability. Corresponding to the whole life cycle of data mining, the combination situation is calculated in the global field: Among them, a(e w ), a(e q ) are two sets of sets e w , e q The global probability of The parallel similarity mining process is described as follows: Among them, sim() is the similarity function, f is the attribute value of the instance data, a f is the probability of occurrence of f, s f is the weight of the data set that has attribute f in common, d f is the correction parameter; Through the above methods, knowledge graph and big data technology are combined to complete data mining and analysis.

Images