CN-121996773-A - Service semantic processing method based on AI and DDD fusion

CN121996773ACN 121996773 ACN121996773 ACN 121996773ACN-121996773-A

Abstract

The application relates to the technical field of semantic processing, and discloses a business semantic processing method based on the fusion of AI and DDD, comprising the steps of obtaining business data; the method comprises the steps of identifying business data to generate structured data, setting a plurality of aggregations according to the structured data, obtaining aggregation association degree of each aggregation, dividing a plurality of limit contexts based on the aggregation association degree, generating a code frame based on the limit contexts, and checking the code frame. The application automatically builds the semantic analysis framework, can comprehensively grasp various semantic information of words and sentences under different conditions, can accurately judge the accurate meaning of each business concept in different contexts, greatly improves the conversion precision of business semantics to technical realization, automatically recognizes semantic divergence under different conditions, can flexibly build a code layer framework according to the existing codes, and effectively improves the analysis efficiency.

Inventors

ZHANG JIALONG
ZHAI YUBO
WEN JIANSONG

Assignees

华能招采数字科技有限公司

Dates

Publication Date: 20260508
Application Date: 20251205

Claims (10)

1. A business semantic processing method based on the fusion of AI and DDD is characterized by comprising the following steps: Acquiring service data; Identifying the service data to generate structured data; Setting a plurality of aggregations according to the structured data; Acquiring an aggregation association degree of each aggregation, and dividing a plurality of limit contexts based on the aggregation association degree; A code frame is generated based on the bounding context, and the code frame is verified.
2. The business semantic processing method based on AI and DDD fusion of claim 1, wherein the generating structured data comprises: identifying semantic units in the business data; Acquiring multidimensional semantic data and position data of each semantic unit; and setting the multidimensional semantic data corresponding to the same semantic unit at different positions as a single semantic information base.
3. The business semantic processing method based on AI and DDD fusion according to claim 2, wherein the setting a plurality of aggregations according to the structured data comprises: selecting semantic units corresponding to each aggregation according to a preset semantic network; Calculating the adaptation degree of each semantic unit; and selecting corresponding multidimensional semantic data from a semantic information base of the semantic unit according to the adaptation degree.
4. The business semantic processing method based on the fusion of AI and DDD according to claim 3, wherein the calculating the fitness of each semantic unit comprises: A semantic element array Y is constructed, y= (Y1, Y2... Yi...yn), wherein Y1 is the first semantic unit in the aggregation; yi is the ith semantic unit in the aggregation, n is the total number of semantic units in the aggregation; Respectively constructing a first vocabulary library for each multidimensional semantic data of the semantic units in the aggregation; Respectively constructing a second vocabulary library for each semantic unit in the aggregation and setting all the second vocabulary libraries in the aggregation as a third vocabulary library; calculating the association adaptation degree of each first vocabulary library and the third vocabulary library of the ith semantic unit respectively; respectively calculating the frequency adaptation degree of each first vocabulary library of the ith semantic unit; Calculating the adaptation degree S according to the associated adaptation degree and the frequency adaptation degree; S=K1*S Switch for closing +K2*S Frequency band ; wherein, K1 is a first coefficient, K2 is a second coefficient, S Switch for closing is a correlation fit, and S Frequency band is a frequency fit.
5. The business semantic processing method based on AI and DDD fusion of claim 4, wherein the calculating the association adaptation degree of each first vocabulary library and the third vocabulary library of the ith semantic unit respectively comprises: Setting a first vocabulary array Ti of Yi, ti= (Ti 1, ti 2..tim..tiq..tiq), wherein, ti1 is a first vocabulary library corresponding to the first multidimensional semantic data of the ith semantic unit in the aggregation; tim is the first vocabulary library corresponding to the mth multidimensional semantic data of the ith semantic unit in the aggregation, q is the number of the first vocabulary libraries corresponding to the ith semantic unit in the aggregation; Respectively setting weight coefficients for the vocabulary in Tim; calculating the occurrence times of each vocabulary in Tim in a third vocabulary library; calculating associated adaptation degree S Switch for closing according to the occurrence times and the weight coefficient; ; Wherein, the For the number of occurrences of the i-th vocabulary, For the i-th vocabulary weight coefficient, p is the total number of vocabularies in Tim.
6. The method for processing business semantics based on AI and DDD fusion of claim 4, wherein the calculating the frequency adaptation degree of each first vocabulary library of the ith semantic unit includes: Acquiring historical service data; calculating the occurrence times of the vocabulary in each first vocabulary library in the historical service data; Acquiring semantic similarity between each first vocabulary library and historical service data; Calculating frequency adaptation S Frequency band according to the semantic similarity and the frequency of occurrence in the historical service data; S Frequency band =k Language words * Calendar with a display ; Wherein, the Calendar with a display For the number of occurrences in the historical traffic data, Is semantic similarity.
7. The method for semantic processing of services based on AI and DDD fusion of claim 6, wherein the obtaining the aggregate relevance of each aggregate comprises: Setting an aggregation number column J, wherein J= (J1, J2., jf., jg., jr), J1 is a first aggregation, J2 is a second aggregation, JF is an f-th aggregation, jg is a g-th aggregation, and Jr is an r-th aggregation; acquiring aggregation root call data among all aggregations; generating a first association degree according to the aggregation root call data; acquiring data transmission data among all aggregations; And generating an aggregate relevance according to the data transmission data and the first relevance.
8. The method for semantic processing of services based on AI and DDD fusion of claim 7, wherein generating a first degree of association from the aggregated root call data comprises: Constructing a calling number sequence Cf of JF to the aggregation root corresponding to other aggregations in the aggregation number sequence J within a preset period, cf= (Cf 1, cf 2..cfg..cfr..cfr), wherein, cf1 is the calling times of JF to the aggregation root corresponding to the first aggregation in J; cf2 is the number of times of the JF to the second aggregation corresponding to the aggregation in J, cfg is the number of times of the JF to the g aggregation corresponding to the aggregation in J, cfr is the number of times of the JF to the r aggregation corresponding to the aggregation in J; Generating a first association G 1 (f, G) between the f-th aggregation and the G-th aggregation; G 1 (f,g)=Cfg/Cfmax; wherein Cfg is the calling number of the aggregation root corresponding to the g aggregation in J, and Cfmax is the maximum value in the array Cf.
9. The method for semantic processing of services based on AI and DDD fusion of claim 8, wherein the generating aggregate relevance comprises: Identifying all entities and value objects in the aggregate Jf; Constructing a calling number sequence Df of JF on other entity and value objects corresponding to aggregation in an aggregation number sequence J within a preset period, wherein Df1 is the calling number of JF on the entity and value object corresponding to the first aggregation in the J, df2 is the calling number of JF on the entity and value object corresponding to the second aggregation in the J, dfg is the calling number of JF on the entity and value object corresponding to the g-th aggregation in the J, and Dfr is the calling number of JF on the entity and value object corresponding to the r-th aggregation in the J; generating a first association correction factor G 2 (f, G) between the f-th aggregation and the G-th aggregation; G 2 (f,g)=1+λ*Dfg/(Cfg+ε); Wherein G 2 (f, G) is a first correlation correction factor between the f-th aggregation and the G-th aggregation, lambda is a first correction intensity, epsilon is a second correction intensity; generating an aggregation association degree G (f, G) between the f-th aggregation and the G-th aggregation; G(f,g)=G 2 （f,g）*G 1 (f,g)。
10. The method for semantic processing of traffic based on AI and DDD fusion of claim 9, wherein the partitioning of the plurality of bounding contexts comprises: setting a first association threshold and a second association threshold; acquiring the aggregation association degree among all aggregations, and setting the aggregation association degree as an association degree matrix; Traversing all the aggregation association degrees in the association degree matrix, and dividing aggregation of which the aggregation association degrees reach a first association degree threshold into the same limit context; And carrying out secondary aggregation division on aggregation meeting the division conditions for aggregation of each undivided limit context.

Description

Service semantic processing method based on AI and DDD fusion Technical Field The application relates to the technical field of semantic processing, in particular to a business semantic processing method based on AI and DDD fusion. Background The business semantic processing is to convert natural language query in the business scene of the enterprise into executable analysis tasks through natural language processing technology, and data-driven intelligent decision support ‌ is realized. Natural language processing technology belongs to a branch of artificial intelligence, and aims to enable a computer to understand and process human language, extract useful information from the language, and help human beings to process various tasks more efficiently. The existing business semantic processing method mainly has the technical defects that firstly, the concept identification of business depends on manual experience, a systematic, comprehensive and automatic semantic analysis mechanism is lacked, meanwhile, the semantics of the same business concept in different business scenes are different, the difference often needs manual identification, quick and accurate identification cannot be automatically carried out, accurate resolution and expression are difficult, secondly, the system architecture design lacks reasonable division, the module division subjectivity is strong, no clearly executable division basis exists, and again, the code generation and business intention matching degree is low, and the rapid organization architecture design by utilizing the existing codes is difficult to flexibly cope with different business demands. Disclosure of Invention In order to solve the technical problems, the application provides a business semantic processing method based on the fusion of AI and DDD, and aims to obtain a technical scheme which can automatically and accurately build a framework, automatically identify semantic divergence under different conditions and flexibly build an organization framework according to the existing codes. In some embodiments of the present application, a service semantic processing method based on AI and DDD fusion is provided, which is characterized in that the method includes: Acquiring service data; Identifying the service data to generate structured data; Setting a plurality of aggregations according to the structured data; Acquiring an aggregation association degree of each aggregation, and dividing a plurality of limit contexts based on the aggregation association degree; A code frame is generated based on the bounding context, and the code frame is verified. In some embodiments of the application, the generating structured data includes: identifying semantic units in the business data; Acquiring multidimensional semantic data and position data of each semantic unit; and setting the multidimensional semantic data corresponding to the same semantic unit at different positions as a single semantic information base. In some embodiments of the application, the setting a plurality of aggregations from the structured data comprises: selecting semantic units corresponding to each aggregation according to a preset semantic network; Calculating the adaptation degree of each semantic unit; and selecting corresponding multidimensional semantic data from a semantic information base of the semantic unit according to the adaptation degree. In some embodiments of the present application, the calculating the fitness of each semantic unit includes: A semantic element array Y is constructed, y= (Y1, Y2... Yi...yn), wherein Y1 is the first semantic unit in the aggregation; yi is the ith semantic unit in the aggregation, n is the total number of semantic units in the aggregation; Respectively constructing a first vocabulary library for each multidimensional semantic data of the semantic units in the aggregation; Respectively constructing a second vocabulary library for each semantic unit in the aggregation and setting all the second vocabulary libraries in the aggregation as a third vocabulary library; calculating the association adaptation degree of each first vocabulary library and the third vocabulary library of the ith semantic unit respectively; respectively calculating the frequency adaptation degree of each first vocabulary library of the ith semantic unit; Calculating the adaptation degree S according to the associated adaptation degree and the frequency adaptation degree; S=K1*S Switch for closing +K2*S Frequency band ; wherein, K1 is a first coefficient, K2 is a second coefficient, S Switch for closing is a correlation fit, and S Frequency band is a frequency fit. In some embodiments of the present application, the calculating the association adaptation degree between each first vocabulary library and the third vocabulary library of the ith semantic unit includes: Setting a first vocabulary array Ti of Yi, ti= (Ti 1, ti 2..tim..tiq..tiq), wherein, ti1 is a first vocabulary library corre