CN113836191B - Intelligent business processing method and system based on big data - Google Patents

Intelligent business processing method and system based on big data Download PDF

Info

Publication number
CN113836191B
CN113836191B CN202110921774.8A CN202110921774A CN113836191B CN 113836191 B CN113836191 B CN 113836191B CN 202110921774 A CN202110921774 A CN 202110921774A CN 113836191 B CN113836191 B CN 113836191B
Authority
CN
China
Prior art keywords
trained
service
business
data
periodic
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202110921774.8A
Other languages
Chinese (zh)
Other versions
CN113836191A (en
Inventor
崔岩莉
高厚良
郝席
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Cic Guoxin Beijing Technology Development Co ltd
Original Assignee
Cic Guoxin Beijing Technology Development Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Cic Guoxin Beijing Technology Development Co ltd filed Critical Cic Guoxin Beijing Technology Development Co ltd
Priority to CN202110921774.8A priority Critical patent/CN113836191B/en
Publication of CN113836191A publication Critical patent/CN113836191A/en
Application granted granted Critical
Publication of CN113836191B publication Critical patent/CN113836191B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
    • G06F16/2465—Query processing support for facilitating data mining operations in structured databases
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • Mathematical Physics (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Computing Systems (AREA)
  • Artificial Intelligence (AREA)
  • Computational Linguistics (AREA)
  • Evolutionary Computation (AREA)
  • Biophysics (AREA)
  • Molecular Biology (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Databases & Information Systems (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Medical Informatics (AREA)
  • Fuzzy Systems (AREA)
  • Probability & Statistics with Applications (AREA)
  • Image Analysis (AREA)

Abstract

The embodiment of the application provides an intelligent service processing method based on big data, firstly, a potential data image corresponding to each stage service event in big data of a service to be processed is obtained, x potential data images are obtained, then, x potential data image shadow degrees are determined according to the x potential data images, x data image key contents are determined according to the x potential data image shadow degrees, and finally, based on the x data image key contents, x stage service demand information corresponding to the big data of the service to be processed is obtained through a machine learning model. By the aid of the method, the machine learning model can be used for accurately and quickly positioning and acquiring the periodic service demand information, other processing modes are not needed for further mining the periodic service demand information, time and resources consumed by mining the periodic service demand information are reduced to a certain extent, and accordingly acquisition efficiency of the periodic service demand information is improved.

Description

Intelligent business processing method and system based on big data
Technical Field
The invention relates to the technical field of big data and intelligent services, in particular to an intelligent service processing method and system based on big data.
Background
Big data is a data set which cannot be captured, managed and processed by a conventional software tool within a certain time range, and is a massive, high-growth-rate and diversified information asset which can have stronger decision-making power, insight discovery power and flow optimization capability only by a new processing mode. With the increase of data volume and the increase of service availability and importance, the landing application of the big data technology is more and more extensive, and the big data technology is mature in the fields of block chain finance, administrative enterprise cloud service, online office, online education and the like. However, with the development of artificial intelligence and cloud computing, the online traffic is increased rapidly, which brings challenges to big data processing and mining, and the related big data mining analysis technology has the problem of low implementation efficiency.
Disclosure of Invention
In order to solve the problems, the invention provides an intelligent business processing method and system based on big data.
In a first aspect, an intelligent business processing method based on big data is provided, which is applied to an intelligent business processing system, and the method includes:
acquiring a potential data image corresponding to each stage business event in to-be-processed business big data to obtain x potential data images, wherein the to-be-processed business big data comprises x stage business events, the stage business events and the potential data images have unique corresponding relations, and x is an integer greater than or equal to 1;
determining x potential data portrait influence degrees according to the x potential data portraits, wherein the potential data portrait influence degrees represent the requirement influence of the periodical service event on the to-be-processed service big data in a first service state, and the potential data portrait influence degrees and the periodical service event have a unique corresponding relation;
determining x data portrait key contents according to the x potential data portrait influence degrees, wherein the data portrait key contents and the staged service events have unique corresponding relations;
based on the x data portrait key contents, acquiring x pieces of staged business demand information corresponding to the business big data to be processed through a machine learning model, wherein the staged business demand information and the staged business events have unique corresponding relations.
Preferably, the obtaining of the potential data portrayal corresponding to each staged business event in the business big data to be processed to obtain x potential data portrayals includes:
aiming at the y-th periodic service event in the service big data to be processed, obtaining a quantitative description corresponding to the y-th periodic service event, wherein the y-th periodic service event is any one of the x periodic service events, and y is an integer which is greater than or equal to 0 and smaller than x;
aiming at the y-th periodic business event in the business big data to be processed, acquiring a potential data portrait label corresponding to the y-th periodic business event;
and aiming at the y-th periodic business event in the business big data to be processed, determining a potential data portrait corresponding to the y-th periodic business event according to a potential data portrait label corresponding to the y-th periodic business event and a quantitative description corresponding to the y-th periodic business event.
Preferably, said determining x potential data representation influence magnitudes from said x potential data representations comprises:
for a y-th periodic service event in the service big data to be processed, determining an event analysis result corresponding to the y-th periodic service event according to a potential data portrait of the y-th periodic service event, wherein the y-th periodic service event is any one of the x periodic service events, and y is an integer greater than or equal to 0 and less than x;
aiming at the y-th periodic business event in the business big data to be processed, acquiring a potential data portrait label corresponding to the y-th periodic business event;
and aiming at the y-th periodic service event in the service big data to be processed, determining the potential data portrait influence degree corresponding to the y-th periodic service event according to the potential data portrait label corresponding to the y-th periodic service event and the event analysis result corresponding to the y-th periodic service event.
Preferably, the determining key content for x data representations based on the x potential data representation influence levels comprises:
aiming at the y-th periodic service event in the service big data to be processed, acquiring a global potential algorithm corresponding to the y-th periodic service event, wherein the y-th periodic service event is any one of the x periodic service events, and y is an integer which is greater than or equal to 0 and smaller than x;
and aiming at the y-th periodic service event in the service big data to be processed, determining the key content of the data image corresponding to the y-th periodic service event according to the global latent algorithm corresponding to the y-th periodic service event and the latent data image influence degree corresponding to the y-th periodic service event.
Preferably, the obtaining x pieces of periodic service demand information corresponding to the service big data to be processed through a machine learning model based on the x pieces of data portrait key content includes:
acquiring x target key contents through at least one group of feature extraction units included in the machine learning model based on the x data portrait key contents;
based on the x target key contents, acquiring x pieces of staged business demand information through at least one group of fully-connected neural networks included in the machine learning model.
Preferably, before obtaining x pieces of periodic service demand information corresponding to the service big data to be processed through a machine learning model based on the x pieces of data portrait key content, the method further includes:
acquiring a potential data image to be trained corresponding to each business event to be trained in business big data to be trained to obtain x potential data images to be trained, wherein the business big data to be trained comprises x business events to be trained, and the business events to be trained and the potential data images to be trained have unique corresponding relations;
determining x potential data image influence degrees to be trained according to the x potential data images to be trained, wherein the potential data image influence degrees to be trained have a unique corresponding relation with the service event to be trained;
determining x data image key contents to be trained according to the x potential data image influence degrees to be trained, wherein the data image key contents to be trained and the business events to be trained have unique corresponding relations;
acquiring x pieces of periodic service demand information to be trained corresponding to the service big data to be trained through a machine learning model to be trained on the basis of the x pieces of data to be trained image key content, wherein the periodic service demand information to be trained and the service event to be trained have a unique corresponding relation;
acquiring x positive example periodic service demand information corresponding to the active service big data;
and training the machine learning model to be trained according to the x regular example periodic service requirement information and the x periodic service requirement information to be trained until the machine learning model meets training evaluation indexes, and outputting the machine learning model.
Preferably, the acquiring x pieces of positive example staged service demand information corresponding to the aggressive service big data includes:
acquiring positive example potential data images corresponding to each positive example service event in the positive example service big data to obtain x positive example potential data images, wherein the positive example service big data comprises x positive example service events, and the positive example service events and the positive example potential data images have unique corresponding relations;
determining x positive example potential data representation influence degrees from the x positive example potential data representations, wherein the positive example potential data representation influence degrees have a unique corresponding relationship with the positive example potential data representations;
determining x pieces of positive example data image key content according to the x pieces of positive example potential data image influence degrees, wherein the positive example data image key content and the positive example service event have a unique corresponding relation;
acquiring x pieces of positive example periodic service demand information corresponding to the positive example service big data through a machine learning model to be trained based on the x pieces of positive example data image key content, wherein the positive example periodic service demand information and the positive example service event have a unique corresponding relation;
the training the machine learning model to be trained according to the x regular example periodic service requirement information and the x periodic service requirement information to be trained until a training evaluation index is met, and outputting the machine learning model includes:
aiming at each business event to be trained in the business big data to be trained and a regular business event corresponding to each business event to be trained, determining the quantization cost between the periodic business requirement information to be trained and the regular business requirement information by adopting a triple loss function, and obtaining x quantization cost;
improving the network parameters of the machine learning model to be trained according to the x quantization costs;
and if the training evaluation index is met, acquiring the machine learning model according to the improved network parameters.
Preferably, before obtaining x pieces of periodic service demand information corresponding to the service big data to be processed through a machine learning model based on the x pieces of data portrait key content, the method further includes:
acquiring a potential data image to be trained corresponding to each business event to be trained in business big data to be trained to obtain x potential data images to be trained, wherein the business big data to be trained comprises x business events to be trained, and the business events to be trained and the potential data images to be trained have unique corresponding relations;
determining x potential data image influence degrees to be trained according to the x potential data images to be trained, wherein the potential data image influence degrees to be trained have a unique corresponding relation with the service event to be trained;
determining x data image key contents to be trained according to the x potential data image influence degrees to be trained, wherein the data image key contents to be trained and the service events to be trained have unique corresponding relations;
acquiring x pieces of periodic service demand information to be trained corresponding to the service big data to be trained through a machine learning model to be trained on the basis of the x pieces of data to be trained image key content, wherein the periodic service demand information to be trained and the service event to be trained have a unique corresponding relation;
acquiring x positive example periodic service demand information corresponding to the active service big data;
acquiring x negative case periodic service demand information corresponding to the big data of the depolarized service;
and training the machine learning model to be trained according to the x positive example periodic service requirement information, the x negative example periodic service requirement information and the x periodic service requirement information to be trained until the training evaluation indexes are met, and outputting the machine learning model.
Preferably, the acquiring x pieces of positive example periodic service demand information corresponding to the aggressive service big data includes:
acquiring positive example potential data images corresponding to each positive example service event in the positive example service big data to obtain x positive example potential data images, wherein the positive example service big data comprises x positive example service events, and the positive example service events and the positive example potential data images have unique corresponding relations;
determining x positive example potential data representation influence degrees from the x positive example potential data representations, wherein the positive example potential data representation influence degrees have a unique corresponding relationship with the positive example potential data representations;
determining x pieces of positive example data image key content according to the x pieces of positive example potential data image influence degrees, wherein the positive example data image key content and the positive example service event have a unique corresponding relation;
acquiring x pieces of positive example periodic service demand information corresponding to the positive example service big data through a machine learning model to be trained based on the x pieces of positive example data image key content, wherein the positive example periodic service demand information and the positive example service event have a unique corresponding relation;
the acquiring x negative case periodic service demand information corresponding to the big data of the depolarized service includes:
acquiring a negative example potential data image corresponding to each negative example service event in the depolarization service big data to obtain x negative example potential data images, wherein the depolarization service big data comprises x negative example service events, and the negative example service events and the negative example potential data images have unique corresponding relations;
determining x negative case potential data representation influence degrees according to the x negative case potential data representations, wherein the negative case potential data representation influence degrees and the negative case potential data representations have unique corresponding relations;
determining x negative case data image key contents according to the influence degree of the x negative case potential data images, wherein the negative case data image key contents and the negative case service events have unique corresponding relations;
acquiring x pieces of negative case periodic service demand information corresponding to the large data of the depolarized service through a machine learning model to be trained on the basis of the x pieces of negative case data image key content, wherein the negative case periodic service demand information and the negative case service event have a unique corresponding relation;
the training the machine learning model to be trained according to the x positive-case periodic service demand information, the x negative-case periodic service demand information and the x periodic service demand information to be trained until a training evaluation index is met, and outputting the machine learning model includes:
aiming at each business event to be trained in the business big data to be trained and a regular business event corresponding to each business event to be trained, determining a first quantization cost between the periodic business requirement information to be trained and the regular business requirement information by adopting a first triple loss function, and obtaining x first quantization costs;
aiming at each business event to be trained and a negative example business event in the business big data to be trained, determining a second quantization cost between the periodic business requirement information to be trained and the negative example periodic business requirement information by adopting a second triple loss function, and obtaining x second quantization costs;
improving the network parameters of the machine learning model to be trained according to the x first quantization costs and the x second quantization costs;
and if the training evaluation index is met, acquiring the machine learning model according to the improved network parameters.
In a second aspect, an intelligent business processing system is provided, which comprises a processor and a memory, which are in communication with each other, wherein the processor is configured to retrieve a computer program from the memory and to implement the method of the first aspect by running the computer program.
In a third aspect, there is provided a computer readable storage medium having stored thereon a computer program which, when executed, implements the method of the first aspect.
The embodiment of the invention provides an intelligent service processing method and system based on big data, and provides an intelligent service processing method based on big data. By the aid of the method, the machine learning model can be used for accurately and quickly positioning and acquiring the periodic service demand information, other processing modes are not needed for further mining the periodic service demand information, time and resources consumed by mining the periodic service demand information are reduced to a certain extent, and accordingly acquisition efficiency of the periodic service demand information is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.
Fig. 1 is a flowchart of an intelligent service processing method based on big data according to an embodiment of the present invention.
Fig. 2 is a block diagram of an intelligent service processing apparatus based on big data according to an embodiment of the present invention.
Detailed Description
In order to better understand the technical solutions of the present invention, the following detailed descriptions of the technical solutions of the present invention are provided with the accompanying drawings and specific embodiments, and it should be understood that the specific features in the embodiments and examples of the present invention are detailed descriptions of the technical solutions of the present invention, and are not limitations of the technical solutions of the present invention, and the technical features in the embodiments and examples of the present invention may be combined with each other without conflict.
In order to improve the technical problems in the background art, the inventor innovatively provides an intelligent business processing method and system based on big data.
Fig. 1 shows an intelligent business processing method based on big data, which is applied to an intelligent business processing system, and the method comprises the following steps S10-S40.
Step S10, obtaining a potential data image corresponding to each periodic service event in the to-be-processed service big data, to obtain x potential data images, where the to-be-processed service big data includes x periodic service events, and the periodic service events and the potential data images have a unique corresponding relationship, and x is an integer greater than or equal to 1.
For example, the areas involved in pending business big data include, but are not limited to, blockchain finance, corporate cloud services, online office, online education, and the like. Accordingly, a staged business event can be understood as an event that needs to be handled and processed in different business scenarios. Further, the potential data representation is used for characterizing potential user requirements or non-significant user requirements of the corresponding periodic business events, such as can be represented in the form of a feature vector or a feature map.
In a related embodiment, the obtaining of the potential data portraits corresponding to each of the staged business events in the to-be-processed business big data to obtain x potential data portraits includes the following contents: aiming at the y-th periodic service event in the service big data to be processed, obtaining a quantitative description (characteristic value) corresponding to the y-th periodic service event, wherein the y-th periodic service event is any one of the x periodic service events, and y is an integer which is greater than or equal to 0 and smaller than x; aiming at the y-th periodic business event in the business big data to be processed, acquiring a potential data portrait label (such as a numerical label) corresponding to the y-th periodic business event; and aiming at the y-th periodic business event in the business big data to be processed, determining a potential data portrait corresponding to the y-th periodic business event according to a potential data portrait label corresponding to the y-th periodic business event and a quantitative description corresponding to the y-th periodic business event. In this way, the potential data representation corresponding to the staged business event can be accurately determined based on the quantization description and the data representation tag, so as to ensure the integrity of the potential data representation.
Step S20, determining x potential data portrait influence degrees according to the x potential data portraits, where the potential data portrait influence degree represents a demand influence of the staged service event on the to-be-processed service big data in the first service state, and the potential data portrait influence degree and the staged service event have a unique corresponding relationship.
It is to be understood that the demand impact may be a demand level or a preference strength. Further, determining x potential data representation influence magnitudes from the x potential data representations may include the following: for a y-th periodic service event in the service big data to be processed, determining an event analysis result (such as a feature splitting result or a feature splitting value) corresponding to the y-th periodic service event according to a potential data representation of the y-th periodic service event, where the y-th periodic service event is any one of the x-th periodic service events, and y is an integer greater than or equal to 0 and smaller than x; aiming at the y-th periodic business event in the business big data to be processed, acquiring a potential data portrait label corresponding to the y-th periodic business event; and aiming at the y-th periodic service event in the service big data to be processed, determining the potential data portrait influence degree corresponding to the y-th periodic service event according to the potential data portrait label corresponding to the y-th periodic service event and the event analysis result corresponding to the y-th periodic service event. By the design, high correlation between the periodical service event and the image influence degree of the data can be ensured.
Step S30, determining x data portrait key contents according to the x potential data portrait influence degrees, wherein the data portrait key contents and the staged service events have unique corresponding relations.
In an embodiment of the application, the data representation key content may be a data representation description or a data representation vector. Further, determining x data portrait key content from the x potential data portrait influence magnitudes includes: aiming at the y-th periodic service event in the service big data to be processed, acquiring a global latent algorithm (a normalization algorithm or a normalization function) corresponding to the y-th periodic service event, wherein the y-th periodic service event is any one of the x periodic service events, and y is an integer which is greater than or equal to 0 and smaller than x; and aiming at the y-th periodic service event in the service big data to be processed, determining the key content of the data image corresponding to the y-th periodic service event according to the global latent algorithm corresponding to the y-th periodic service event and the latent data image influence degree corresponding to the y-th periodic service event. Therefore, by carrying out global processing, the service adaptability and the data format consistency of the key contents of the data representation corresponding to different staged service events can be ensured, and the error in the subsequent required information analysis can be reduced.
Step S40, based on the x data portrait key contents, acquiring x pieces of staged business demand information corresponding to the to-be-processed business big data through a machine learning model, wherein the staged business demand information and the staged business events have unique corresponding relations.
In the embodiment of the application, the periodic service demand information can be service demands or service demands at different stages, and optimization and upgrade of service products can be realized by determining the periodic service demand information, so that targeted service during subsequent service interaction is ensured, and data quality of subsequently obtained service big data is ensured.
In some possible embodiments, based on the x pieces of data portrait key content, acquiring x pieces of staged business demand information corresponding to the to-be-processed business big data through a machine learning model, including: acquiring x target key contents through at least one group of feature extraction units included in the machine learning model based on the x data portrait key contents; based on the x target key contents, acquiring x pieces of staged business demand information through at least one group of fully-connected neural networks included in the machine learning model.
In some optional embodiments, before implementing step S40, the method further comprises: acquiring a potential data image to be trained corresponding to each business event to be trained in business big data to be trained to obtain x potential data images to be trained, wherein the business big data to be trained comprises x business events to be trained, and the business events to be trained and the potential data images to be trained have unique corresponding relations; determining x potential data image influence degrees to be trained according to the x potential data images to be trained, wherein the potential data image influence degrees to be trained have a unique corresponding relation with the service event to be trained; determining x data image key contents to be trained according to the x potential data image influence degrees to be trained, wherein the data image key contents to be trained and the business events to be trained have unique corresponding relations; acquiring x pieces of periodic service demand information to be trained corresponding to the service big data to be trained through a machine learning model to be trained on the basis of the x pieces of data to be trained image key content, wherein the periodic service demand information to be trained and the service event to be trained have a unique corresponding relation; acquiring x positive example periodic service demand information corresponding to the active service big data; and training the machine learning model to be trained according to the x regular example periodic service requirement information and the x periodic service requirement information to be trained until the machine learning model meets training evaluation indexes, and outputting the machine learning model.
In the embodiment of the present application, the aggressive traffic big data may be understood as a positive sample. Further, acquiring x positive example periodic service demand information corresponding to the big data of the aggressive service, including: acquiring positive example potential data images corresponding to each positive example service event in the positive example service big data to obtain x positive example potential data images, wherein the positive example service big data comprises x positive example service events, and the positive example service events and the positive example potential data images have unique corresponding relations; determining x positive example potential data representation influence degrees from the x positive example potential data representations, wherein the positive example potential data representation influence degrees have a unique corresponding relationship with the positive example potential data representations; determining x pieces of positive example data image key content according to the x pieces of positive example potential data image influence degrees, wherein the positive example data image key content and the positive example service event have a unique corresponding relation; based on the x pieces of positive case data image key content, x pieces of positive case staged business requirement information corresponding to the positive case business big data are obtained through a machine learning model to be trained, wherein the positive case staged business requirement information and the positive case business events have unique corresponding relations.
On the basis of the above contents, training the machine learning model to be trained according to the x regular example periodic service requirement information and the x periodic service requirement information to be trained until a training evaluation index is met, and outputting the machine learning model, including: aiming at each business event to be trained in the business big data to be trained and a regular business event corresponding to each business event to be trained, determining the quantization cost between the periodic business requirement information to be trained and the regular business requirement information by adopting a triple loss function, and obtaining x quantization cost; improving the network parameters of the machine learning model to be trained according to the x quantization costs; and if the training evaluation index is met, acquiring the machine learning model according to the improved network parameters.
In the embodiment of the present application, the quantization cost may be a loss function.
It can be understood that by introducing the positive examples to perform model training, the machine learning model can be ensured to be capable of accurately positioning and excavating business requirement information in subsequent use.
On the basis of the above contents, countertraining can be performed by introducing negative examples, so that the classification performance and the anti-interference performance of the machine learning model are further improved. Based on this, before implementing step S40, the following may be included: acquiring a potential data image to be trained corresponding to each business event to be trained in business big data to be trained to obtain x potential data images to be trained, wherein the business big data to be trained comprises x business events to be trained, and the business events to be trained and the potential data images to be trained have unique corresponding relations; determining x potential data image influence degrees to be trained according to the x potential data images to be trained, wherein the potential data image influence degrees to be trained have a unique corresponding relation with the service event to be trained; determining x data image key contents to be trained according to the x potential data image influence degrees to be trained, wherein the data image key contents to be trained and the business events to be trained have unique corresponding relations; acquiring x pieces of periodic service demand information to be trained corresponding to the service big data to be trained through a machine learning model to be trained on the basis of the x pieces of data to be trained image key content, wherein the periodic service demand information to be trained and the service event to be trained have a unique corresponding relation; acquiring x positive example periodic service demand information corresponding to the active service big data; acquiring x negative case periodic service demand information corresponding to the big data of the depolarized service; and training the machine learning model to be trained according to the x positive example periodic service requirement information, the x negative example periodic service requirement information and the x periodic service requirement information to be trained until the machine learning model meets training evaluation indexes, and outputting the machine learning model.
On the basis of the above contents, acquiring x positive example staged service demand information corresponding to the aggressive service big data, including: acquiring positive example potential data images corresponding to each positive example service event in the positive example service big data to obtain x positive example potential data images, wherein the positive example service big data comprises x positive example service events, and the positive example service events and the positive example potential data images have unique corresponding relations; determining x positive example potential data representation influence degrees from the x positive example potential data representations, wherein the positive example potential data representation influence degrees have a unique corresponding relationship with the positive example potential data representations; determining x pieces of positive example data image key content according to the x pieces of positive example potential data image influence degrees, wherein the positive example data image key content and the positive example service event have a unique corresponding relation; based on the x pieces of positive case data image key content, x pieces of positive case staged business requirement information corresponding to the positive case business big data are obtained through a machine learning model to be trained, wherein the positive case staged business requirement information and the positive case business events have unique corresponding relations.
Optionally, the obtaining x pieces of negative case periodic service demand information corresponding to the big data of the depolarizing service includes: acquiring a negative example potential data image corresponding to each negative example service event in the depolarization service big data to obtain x negative example potential data images, wherein the depolarization service big data comprises x negative example service events, and the negative example service events and the negative example potential data images have unique corresponding relations; determining x negative case potential data representation influence degrees according to the x negative case potential data representations, wherein the negative case potential data representation influence degrees and the negative case potential data representations have unique corresponding relations; determining x negative case data image key contents according to the influence degree of the x negative case potential data images, wherein the negative case data image key contents and the negative case service events have unique corresponding relations; based on the x negative example data image key contents, acquiring x negative example periodic service requirement information corresponding to the large data of the depolarizing service through a machine learning model to be trained, wherein the negative example periodic service requirement information and the negative example service event have a unique corresponding relation.
On the basis of the above contents, training the machine learning model to be trained according to the x positive example periodic service requirement information, the x negative example periodic service requirement information, and the x periodic service requirement information to be trained until a training evaluation index is met, and outputting the machine learning model, including: aiming at each business event to be trained in the business big data to be trained and a regular business event corresponding to each business event to be trained, determining a first quantization cost between the periodic business requirement information to be trained and the regular business requirement information by adopting a first triple loss function, and obtaining x first quantization costs; aiming at each business event to be trained and a negative example business event in the business big data to be trained, determining a second quantization cost between the periodic business requirement information to be trained and the negative example periodic business requirement information by adopting a second triple loss function, and obtaining x second quantization costs; improving the network parameters of the machine learning model to be trained according to the x first quantization costs and the x second quantization costs; and if the training evaluation index is met, acquiring the machine learning model according to the improved network parameters.
In the above embodiment, the periodic service demand information can also be understood as local demand information, and by performing local analysis, real-time monitoring processing on different big data services can be realized, and massive and diverse service data and service information can be processed in time, so that timeliness demand and accuracy demand of service big data analysis can be met.
In the embodiment of the application, various intelligent services are monitored by utilizing big data mining and artificial intelligence technologies, and big data analysis is carried out on mass data generated in the process of handling related services, so that the real-time service demand and the data quality of the data are ensured.
On the basis, please refer to fig. 2 in combination, there is provided an intelligent business processing apparatus 200 based on big data, which is applied to an intelligent business processing system, and the apparatus includes:
the image obtaining module 210 is configured to obtain a potential data image corresponding to each periodic service event in to-be-processed service big data to obtain x potential data images, where the to-be-processed service big data includes x periodic service events, the periodic service events and the potential data images have a unique corresponding relationship, and x is an integer greater than or equal to 1;
an influence analysis module 220, configured to determine x potential data portrait influence degrees according to the x potential data portraits, where the potential data portrait influence degree represents a requirement influence of the periodic service event on the to-be-processed service big data in a first service state, and the potential data portrait influence degree and the periodic service event have a unique corresponding relationship;
a content determining module 230, configured to determine x number of key data portrait contents according to the x number of potential data portrait influence degrees, where the key data portrait contents and the periodic service event have a unique corresponding relationship;
and the requirement mining module 240 is configured to obtain x pieces of periodic service requirement information corresponding to the to-be-processed service big data through a machine learning model based on the x pieces of data portrait key content, where the periodic service requirement information and the periodic service event have a unique corresponding relationship.
For the description of the functional modules, reference may be made to the description of the method shown in fig. 1, which is not described herein again.
On the basis of the above, an intelligent business processing system is provided, which comprises a processor and a memory, which are communicated with each other, wherein the processor is used for retrieving a computer program from the memory and implementing the method by running the computer program.
On the basis of the above, a computer-readable storage medium is provided, on which a computer program is stored, which computer program realizes the above-described method when executed.
In summary, based on the above scheme, an intelligent service processing method based on big data is provided, which includes obtaining potential data images corresponding to each staged service event in big data of a service to be processed, obtaining x potential data images, determining x potential data image influence degrees according to the x potential data images, determining x data image key contents according to the x potential data image influence degrees, and finally obtaining x staged service demand information corresponding to the big data of the service to be processed through a machine learning model based on the x data image key contents. By the aid of the method, the machine learning model can be used for accurately and quickly positioning and acquiring the periodic service demand information, other processing modes are not needed for further mining the periodic service demand information, time and resources consumed by mining the periodic service demand information are reduced to a certain extent, and accordingly acquisition efficiency of the periodic service demand information is improved.
The above are merely examples of the present application and are not intended to limit the present application. Various modifications and changes may occur to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims (6)

1. An intelligent business processing method based on big data is applied to an intelligent business processing system, and the method comprises the following steps:
acquiring a potential data image corresponding to each stage business event in to-be-processed business big data to obtain x potential data images, wherein the to-be-processed business big data comprises x stage business events, the stage business events and the potential data images have unique corresponding relations, and x is an integer greater than or equal to 1;
determining x potential data portrait influence degrees according to the x potential data portraits, wherein the potential data portrait influence degrees represent the requirement influence of the periodical service event on the to-be-processed service big data in a first service state, and the potential data portrait influence degrees and the periodical service event have a unique corresponding relation;
determining x data portrait key contents according to the x potential data portrait influence degrees, wherein the data portrait key contents and the staged service events have unique corresponding relations;
acquiring x pieces of staged business demand information corresponding to the business big data to be processed through a machine learning model based on the x pieces of data portrait key content, wherein the staged business demand information and the staged business events have unique corresponding relations;
the method for obtaining the potential data portraits corresponding to each stage business event in the business big data to be processed to obtain x potential data portraits includes the following steps:
aiming at the y-th periodic service event in the service big data to be processed, obtaining a quantitative description corresponding to the y-th periodic service event, wherein the y-th periodic service event is any one of the x periodic service events, and y is an integer which is greater than or equal to 0 and smaller than x;
aiming at the y-th periodic business event in the business big data to be processed, acquiring a potential data portrait label corresponding to the y-th periodic business event;
aiming at the y-th periodic business event in the business big data to be processed, determining a potential data image corresponding to the y-th periodic business event according to a potential data image label corresponding to the y-th periodic business event and a quantitative description corresponding to the y-th periodic business event;
wherein said determining x potential data representation influence magnitudes from said x potential data representations comprises:
for a y-th periodic service event in the service big data to be processed, determining an event analysis result corresponding to the y-th periodic service event according to a potential data portrait of the y-th periodic service event, wherein the y-th periodic service event is any one of the x periodic service events, and y is an integer greater than or equal to 0 and less than x;
aiming at the y-th periodic business event in the business big data to be processed, acquiring a potential data portrait label corresponding to the y-th periodic business event;
aiming at the y-th periodic service event in the service big data to be processed, determining the potential data portrait influence degree corresponding to the y-th periodic service event according to the potential data portrait label corresponding to the y-th periodic service event and the event analysis result corresponding to the y-th periodic service event;
wherein said determining x data portrait key content from said x potential data portrait influence magnitudes comprises:
aiming at the y-th periodic business event in the business big data to be processed, acquiring a global latent algorithm corresponding to the y-th periodic business event, wherein the y-th periodic business event is any one of the x periodic business events, and y is an integer which is greater than or equal to 0 and smaller than x;
for the y-th periodic service event in the service big data to be processed, determining the key content of the data image corresponding to the y-th periodic service event according to the global latent algorithm corresponding to the y-th periodic service event and the latent data image influence degree corresponding to the y-th periodic service event;
acquiring x pieces of staged service demand information corresponding to the service big data to be processed through a machine learning model based on the x pieces of data portrait key content, wherein the acquiring comprises the following steps:
acquiring x target key contents through at least one group of feature extraction units included in the machine learning model based on the x data portrait key contents;
based on the x target key contents, acquiring x pieces of staged business demand information through at least one group of fully-connected neural networks included in the machine learning model.
2. The method according to claim 1, wherein before acquiring x pieces of staged business demand information corresponding to the big data of the business to be processed through a machine learning model based on the x pieces of data portrait key content, the method further comprises:
acquiring a potential data image to be trained corresponding to each business event to be trained in business big data to be trained to obtain x potential data images to be trained, wherein the business big data to be trained comprises x business events to be trained, and the business events to be trained and the potential data images to be trained have unique corresponding relations;
determining x potential data image influence degrees to be trained according to the x potential data images to be trained, wherein the potential data image influence degrees to be trained have a unique corresponding relation with the service event to be trained;
determining x data image key contents to be trained according to the x potential data image influence degrees to be trained, wherein the data image key contents to be trained and the business events to be trained have unique corresponding relations;
acquiring x pieces of periodic service demand information to be trained corresponding to the service big data to be trained through a machine learning model to be trained on the basis of the x pieces of data to be trained image key content, wherein the periodic service demand information to be trained and the service event to be trained have a unique corresponding relation;
acquiring x positive example periodic service demand information corresponding to the active service big data;
and training the machine learning model to be trained according to the x regular example periodic service requirement information and the x periodic service requirement information to be trained until the machine learning model meets training evaluation indexes, and outputting the machine learning model.
3. The method according to claim 2, wherein the obtaining x positive example periodic service requirement information corresponding to the aggressive service big data includes:
acquiring positive example potential data images corresponding to each positive example service event in the positive example service big data to obtain x positive example potential data images, wherein the positive example service big data comprises x positive example service events, and the positive example service events and the positive example potential data images have unique corresponding relations;
determining x positive example potential data representation influence degrees from the x positive example potential data representations, wherein the positive example potential data representation influence degrees have a unique corresponding relationship with the positive example potential data representations;
determining x pieces of positive example data image key content according to the x pieces of positive example potential data image influence degrees, wherein the positive example data image key content and the positive example service event have a unique corresponding relation;
acquiring x pieces of positive example periodic service demand information corresponding to the positive example service big data through a machine learning model to be trained based on the x pieces of positive example data image key content, wherein the positive example periodic service demand information and the positive example service event have a unique corresponding relation;
the training the machine learning model to be trained according to the x regular example periodic service requirement information and the x periodic service requirement information to be trained until a training evaluation index is met, and outputting the machine learning model includes:
aiming at each business event to be trained in the business big data to be trained and a regular business event corresponding to each business event to be trained, determining the quantization cost between the periodic business requirement information to be trained and the regular business requirement information by adopting a triple loss function, and obtaining x quantization cost;
improving the network parameters of the machine learning model to be trained according to the x quantization costs;
and if the training evaluation index is met, acquiring the machine learning model according to the improved network parameters.
4. The method according to claim 1, wherein before acquiring x pieces of staged business demand information corresponding to the big data of the business to be processed through a machine learning model based on the x pieces of data portrait key content, the method further comprises:
acquiring a potential data image to be trained corresponding to each business event to be trained in business big data to be trained to obtain x potential data images to be trained, wherein the business big data to be trained comprises x business events to be trained, and the business events to be trained and the potential data images to be trained have unique corresponding relations;
determining x potential data image influence degrees to be trained according to the x potential data images to be trained, wherein the potential data image influence degrees to be trained have a unique corresponding relation with the service event to be trained;
determining x data image key contents to be trained according to the x potential data image influence degrees to be trained, wherein the data image key contents to be trained and the business events to be trained have unique corresponding relations;
acquiring x pieces of periodic service demand information to be trained corresponding to the service big data to be trained through a machine learning model to be trained on the basis of the x pieces of data to be trained image key content, wherein the periodic service demand information to be trained and the service event to be trained have a unique corresponding relation;
acquiring x positive example periodic service demand information corresponding to the active service big data;
acquiring x negative case periodic service demand information corresponding to the big data of the depolarized service;
and training the machine learning model to be trained according to the x positive example periodic service requirement information, the x negative example periodic service requirement information and the x periodic service requirement information to be trained until the machine learning model meets training evaluation indexes, and outputting the machine learning model.
5. The method according to claim 4, wherein the obtaining x positive example periodic service requirement information corresponding to the aggressive service big data includes:
acquiring positive example potential data images corresponding to each positive example service event in the positive example service big data to obtain x positive example potential data images, wherein the positive example service big data comprises x positive example service events, and the positive example service events and the positive example potential data images have unique corresponding relations;
determining x positive example potential data representation influence degrees from the x positive example potential data representations, wherein the positive example potential data representation influence degrees have a unique corresponding relationship with the positive example potential data representations;
determining x pieces of positive example data image key content according to the x pieces of positive example potential data image influence degrees, wherein the positive example data image key content and the positive example service event have a unique corresponding relation;
based on the x pieces of positive case data image key content, acquiring x pieces of positive case periodic business requirement information corresponding to the positive case business big data through a machine learning model to be trained, wherein the positive case periodic business requirement information and the positive case business events have unique corresponding relation;
the acquiring x negative case periodic service demand information corresponding to the big data of the depolarized service includes:
acquiring a negative example potential data image corresponding to each negative example service event in the depolarization service big data to obtain x negative example potential data images, wherein the depolarization service big data comprises x negative example service events, and the negative example service events and the negative example potential data images have unique corresponding relations;
determining x negative case potential data representation influence degrees according to the x negative case potential data representations, wherein the negative case potential data representation influence degrees and the negative case potential data representations have unique corresponding relations;
determining x negative case data image key contents according to the influence degree of the x negative case potential data images, wherein the negative case data image key contents and the negative case service events have unique corresponding relations;
acquiring x pieces of negative case periodic service demand information corresponding to the large data of the depolarized service through a machine learning model to be trained on the basis of the x pieces of negative case data image key content, wherein the negative case periodic service demand information and the negative case service event have a unique corresponding relation;
the training the machine learning model to be trained according to the x positive-case periodic service demand information, the x negative-case periodic service demand information and the x periodic service demand information to be trained until a training evaluation index is met, and outputting the machine learning model includes:
aiming at each business event to be trained in the business big data to be trained and a regular business event corresponding to each business event to be trained, determining a first quantization cost between the periodic business requirement information to be trained and the regular business requirement information by adopting a first triple loss function, and obtaining x first quantization costs;
aiming at each business event to be trained and a negative example business event in the business big data to be trained, determining a second quantization cost between the periodic business requirement information to be trained and the negative example periodic business requirement information by adopting a second triple loss function, and obtaining x second quantization costs;
improving the network parameters of the machine learning model to be trained according to the x first quantization costs and the x second quantization costs;
and if the training evaluation index is met, acquiring the machine learning model according to the improved network parameters.
6. An intelligent business processing system comprising a processor and a memory in communication with each other, the processor being configured to retrieve a computer program from the memory and to implement the method of any of claims 1-5 by running the computer program.
CN202110921774.8A 2021-08-12 2021-08-12 Intelligent business processing method and system based on big data Active CN113836191B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110921774.8A CN113836191B (en) 2021-08-12 2021-08-12 Intelligent business processing method and system based on big data

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110921774.8A CN113836191B (en) 2021-08-12 2021-08-12 Intelligent business processing method and system based on big data

Publications (2)

Publication Number Publication Date
CN113836191A CN113836191A (en) 2021-12-24
CN113836191B true CN113836191B (en) 2022-08-02

Family

ID=78963274

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110921774.8A Active CN113836191B (en) 2021-08-12 2021-08-12 Intelligent business processing method and system based on big data

Country Status (1)

Country Link
CN (1) CN113836191B (en)

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11068916B2 (en) * 2017-06-26 2021-07-20 Kronos Technology Systems Limited Partnershi Using machine learning to predict retail business volume
CN112840363A (en) * 2018-09-11 2021-05-25 格林伊登美国控股有限责任公司 Method and system for predicting workload demand in customer travel applications
CN112015962A (en) * 2020-07-24 2020-12-01 北京艾巴斯智能科技发展有限公司 Government affair intelligent big data center system architecture
CN113536127A (en) * 2021-01-12 2021-10-22 陈漩 Data processing method based on big data and artificial intelligence and cloud server
CN112839102A (en) * 2021-02-04 2021-05-25 葛天齐 Business processing method applied to big data image pushing and machine learning server

Also Published As

Publication number Publication date
CN113836191A (en) 2021-12-24

Similar Documents

Publication Publication Date Title
CN110782042B (en) Method, device, equipment and medium for combining horizontal federation and vertical federation
CN111984383B (en) Service data processing method and cloud platform based on cloud network fusion and artificial intelligence
Gagné et al. Distributed beagle: An environment for parallel and distributed evolutionary computations
EP3785128A2 (en) System and method for creating recommendation of splitting and merging microservice
Ren et al. A novel deep learning method for application identification in wireless network
CN114611081B (en) Account type identification method, device, equipment, storage medium and product
CN117271100B (en) Algorithm chip cluster scheduling method, device, computer equipment and storage medium
CN111949720B (en) Data analysis method based on big data and artificial intelligence and cloud data server
CN113836191B (en) Intelligent business processing method and system based on big data
CN113792892A (en) Federal learning modeling optimization method, apparatus, readable storage medium, and program product
CN110826867B (en) Vehicle management method, device, computer equipment and storage medium
WO2023035526A1 (en) Object sorting method, related device, and medium
CN116109102A (en) Resource allocation method and system based on genetic algorithm
CN115712763A (en) Network data processing method and platform based on artificial intelligence and big data
US20230297885A1 (en) Big data-based modular ai engine server and driving method of the same
CN114329127A (en) Characteristic box dividing method, device and storage medium
CN115983392A (en) Method, device, medium and electronic device for determining quantum program mapping relation
CN113010288A (en) Scheduling method and device of cloud resources and computer storage medium
CN113535540A (en) Data testing method, device, system, equipment and storage medium
Thomas et al. Understanding redundancy in discrete multi-agent communication
Zakutynskyi Finding the Optimal Number of Computing Containers in IoT Systems: Application of Mathematical Modeling Methods
CN111917854B (en) Cooperation type migration decision method and system facing MCC
CN113946758B (en) Data identification method, device, equipment and readable storage medium
CN116957067B (en) Reinforced federal learning method and device for public safety event prediction model
CN109995756B (en) Online single-classification active machine learning method for information system intrusion detection

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
CB03 Change of inventor or designer information

Inventor after: Cui Yanli

Inventor after: Gao Houliang

Inventor after: Hao Xi

Inventor before: Cui Yanli

Inventor before: Hao Xi

Inventor before: Gao Houliang

CB03 Change of inventor or designer information
GR01 Patent grant
GR01 Patent grant