CN115048080A - Science and technology special distributor service cloud platform construction system based on big data - Google Patents

Abstract

The invention discloses a technology special distributor service cloud platform construction system based on big data, which constructs application systems such as a technology special distributor service cloud platform service entrance (a portal website, a comet-nong-letter APP and a comet-nong-letter applet), a technology special distributor registration management and service effect evaluation system, a comet-nong-letter service system, an agricultural resource management system, a technology special distributor data analysis display system and a service cloud platform management system by using new-generation information technologies such as mobile internet, cloud computing, big data, internet of things and artificial intelligence, and improves the overall informatization level of the technology special distributor work by constructing a platform.

Description

Science and technology special distributor service cloud platform construction system based on big data

Technical Field

The invention relates to the field related to a big data science and technology dispatcher service cloud platform, in particular to a science and technology dispatcher service cloud platform construction system based on big data.

Background

In order to further improve the informatization management level of the science and technology specially-assigned personnel, establish a perfect science and technology specially-assigned personnel assessment index system and an exit mechanism, carry out dynamic management, strengthen dynamic monitoring on the work of the science and technology specially-assigned personnel and perfect the statistical report work of the science and technology specially-assigned personnel, the informatization management measures for selecting and managing the science and technology specially-assigned personnel in the province need to be further perfected, and a special science and technology specially-assigned personnel service cloud platform construction system needs to be used.

The existing platform system has poor service effect on the science and technology specializer during actual use, so that the efficiency and the effect are lower when the accurate service, the remote service and the dynamic service are carried out to provide the informatization support, the whole informatization level is poor, and the informatization management measures on the science and technology specializer are fewer.

Disclosure of Invention

Therefore, in order to solve the above disadvantages, the invention provides a technology dispatcher service cloud platform construction system based on big data.

In order to achieve the purpose, the invention adopts the following technical scheme: a science and technology special dispatcher service cloud platform construction system based on big data is characterized in that a system target architecture is divided into six layers: the system comprises an infrastructure layer, a data layer, an application supporting layer, an application layer, an access layer, a display layer and a science and technology special distributor supply and demand docking system;

and (3) displaying layer: the display layer is an entrance of the platform facing to the final user, is a main entrance and an interactive interface for various users to obtain required services, and realizes login of user access, integrated access of various business applications, personalized display, user and safety management and other aspects, and is developed based on html5, css, vue.js and G6 languages and a technical framework; the display layer is mainly applied as follows: PC end page, WeChat applet, APP (including Android version and iOS version);

an access layer: the access layer is a communication service layer between a system front-end system and a system back-end system, nginx is used as reverse proxy service, the presentation layer and the access layer communicate by using https protocol, and the access layer provides an html page and a respAPI interface for the presentation layer;

an application layer: the application layer is built based on a SprigCloud system architecture, platform service logic is realized, cross-platform and cross-database applications are supported, and the application layer service applications are mainly divided into a science and technology special dispatcher registration management and service effect evaluation system, a science and technology special dispatcher data analysis and display system, a coment-rural communication service system, an agricultural technology resource management system and a service cloud platform management system;

applying a support layer: the application support layer realizes the intercommunication between the infrastructure layer and the application layer by utilizing various universal platforms and provides uniform identity authentication for service application; the main units of the application supporting layer are as follows: the system comprises a password application, a workflow engine, a search engine, a message service engine, a file service, unified log management, unified identity authentication and capability service integration;

and (3) a data layer: the data layer is established on hardware and an operating system and is mainly used for storing basic data of the platform and providing data storage support for the whole application, and the data storage of the project is mainly an Oracle database and assisted by a non-relational database, an elastic search and a memory database redis; the main units of the data layer are as follows: the system comprises a scientific and technological special distributor information database, a comet-nong-letter service information database, a public service information database, a basic information database, a data storage and calculation system, a data acquisition and aggregation and data analysis system;

infrastructure layer: the infrastructure layer comprises a system layer and a network layer, is a foundation for supporting the operation of the information system, comprises hardware equipment and a cloud platform, and is a physical environment for bearing the whole information system; the infrastructure layer is provided by the government cloud. (ii) a

Science and technology special distributor supply and demand docking system: the method comprises the steps of extracting topics of personal information of a science and technology specializer and demand content of a service object by using an LDA model, respectively obtaining probability distribution of the topics, measuring topic similarity of the science and technology specializer and user demands x and y through Jensen-Shannon divergence, carrying out demand matching, and carrying out sequencing recommendation according to the similarity.

Preferably, the system for registration management and service performance evaluation of the scientific and technological speciality mainly comprises: the system comprises a workbench, a kort registration, a basic information management, a kort declaration, a work report, a requirement butt joint, a kort management, an audit management, an expert information, an expert studio, a kort wind collecting management, a kort evaluation and assessment, a unit information, a declaration plan management, a kort declaration module management, a work report template management, a data analysis, a service achievement integral calculation and a unit management;

the scientific and technological speciality data analysis and display system mainly comprises: the method comprises the following steps of performing analysis on a person assigned by a department, performing analysis on information of the department, performing analysis on a department composition, performing analysis on service distribution of the department, performing analysis on poverty-relief county, performing analysis on assistance effect of the department, performing analysis on city department distribution and performing a department map;

the comet-nong-letter service system mainly comprises: user registration, consultation watching, expert inquiring, agricultural video, requirement docking, remote consultation, expert mobile studio, personal center, APP application setting, content auditing, service data management and agricultural video management;

the agricultural technology resource management system mainly comprises: agricultural knowledge management, agricultural knowledge auditing, agricultural picture management, agricultural picture auditing, agricultural training video management, agricultural skill training video auditing, identification management and classification management;

the service cloud platform management system mainly comprises: questionnaire management, content management, user management, log management, user feedback, data dictionary management, external service support, and security settings.

Preferably, the application support layer mainly comprises the following types:

micro-service application support: service administration, service link tracing (skywalk), service log collection and analysis based on the SpringCloud micro-service framework;

a middleware: a workflow engine, a report engine, a search engine and a message service engine;

the business application class supports applications: the method comprises the steps of uniform identity authentication, password application, file service, protocol analysis, uniform log management and capability service integration;

accessing a third-party interface: the system comprises a government affair data sharing center interface, an Aliyun object storage product interface, an Aliyun video on demand product interface, an Aliyun sensitive word filtering product interface, an Tencent real-time audio and video interface, an Tencent short message service interface, an Tencent cloud map open platform interface and a Gagde map open platform interface.

Preferably, the infrastructure layer built-in hardware device is: server equipment, storage equipment, network equipment and security equipment; the cloud platform is as follows: computing resource pools, storage resource pools, network resource pools, security resource pools, server virtualization, network virtualization, and storage virtualization.

Preferably, the application layer is built based on a SprigCloud system architecture and supports cross-platform and cross-database applications;

the method provides a simple development mode for configuration management, service administration, circuit breakers, intelligent routing, micro-services, control buses, global locks, decision election, distributed sessions and cluster state management operations related to a micro-service architecture.

Preferably, the application layer builds a built-in SpringCloud access gateway zuul, service communication, a business engine, service registration and discovery Eureka and a configuration center Config based on a SprigCloud system architecture.

Preferably, the system interfaces with an external network platform through an external server interface/capability open service.

Preferably, the network platform mainly comprises: the system comprises a weather station, a government affair information sharing platform, an Ali cloud platform, an Tengcong cloud platform and a rural scientific and technological information resource sharing and service platform.

Preferably, the pairing algorithm of the science and technology speciality supply and demand docking system is as follows:

wherein

Is the subject distribution of x and y described by the technology specializer and user requirements,

is the distribution Jensen-Shannon divergence with the value range of [0, 1%]The larger the value is, the higher the degree of dispersion is, and the calculation formula is as follows:

KL (Kullback-Liebler divergence) divergence was:

the invention has the beneficial effects that:

according to the invention, by applying new generation information technologies such as mobile internet, cloud computing, big data, internet of things and artificial intelligence, application systems such as a scientific and technological dispatcher service cloud platform service entrance (a portal website, a comet-agriculture-trust APP and a comet-agriculture-trust applet), a scientific and technological dispatcher registration management and service effect evaluation system, a comet-agriculture-trust service system, an agricultural technology resource management system, a scientific and technological dispatcher data analysis and display system and a service cloud platform management system are built through perfection, optimization and innovation, and the platform is built to provide informatization support for the scientific and technological dispatcher to develop accurate service, remote service and dynamic service, so that the overall informatization level of work of the scientific and technological dispatcher is improved.

Drawings

FIG. 1 is a schematic diagram of the system architecture of the present invention;

FIG. 2 is a schematic diagram of a technical architecture of the present invention;

FIG. 3 is a schematic diagram of the operating principle of the eureka of the present invention;

FIG. 4 is a schematic diagram of the operating principle of the configuration center of the present invention;

FIG. 5 is a schematic diagram of the message bus operation of the present invention;

FIG. 6 is a schematic diagram of an Tencent real-time audio/video product architecture according to the present invention;

FIG. 7 is a general flow diagram of a real-time voice-video call in accordance with the present invention;

FIG. 8 is an expert online schematic of the present invention;

FIG. 9 is a schematic diagram of the expert of the present invention not being online;

FIG. 10 is a diagram of an expert offline with a caller hanging up in accordance with the present invention;

FIG. 11 is an expert online illustration of the invention when a calling party hangs up;

FIG. 12 is a schematic diagram of the called party active hangup of the present invention;

FIG. 13 is a schematic view of the overall business process of the Aliyun AXB privacy intermediary number of the present invention;

FIG. 14 is a schematic diagram of the invention for obtaining privacy numbers;

FIG. 15 is a schematic diagram of the dialing of the present invention;

FIG. 16 is a schematic diagram of the present invention for canceling dialing;

fig. 17 is a schematic diagram of the end of call of the present invention.

Detailed Description

In order to further explain the technical solution of the present invention, the following detailed description is given by way of specific examples.

The system target architecture is divided into six layers: the system comprises an infrastructure layer, a data layer, an application supporting layer, an application layer, an access layer, a display layer and a science and technology special distributor supply and demand docking system;

a display layer: the display layer is an entrance of the platform facing to the final user, is a main entrance and an interactive interface for various users to obtain required services, and realizes login of user access, integrated access of various business applications, personalized display, user and safety management and other aspects, and is developed based on html5, css, vue.js and G6 languages and a technical framework; the display layer is mainly applied as follows: PC end page (portal site), WeChat applet, APP (including Android version and iOS version);

science and technology specializer service cloud platform portal:

a scientific and technological specializer service cloud platform portal system is built to widen channels for farmers to obtain agricultural technical information resources, improve the service efficiency of scientific and technological specializer and improve the utilization rate of agricultural information resources, and the portal system mainly comprises contents such as notice bulletins, scientific and technological specializer trends, agricultural technical resources, online service, supply and demand butt joint, scientific and technological wind collection, service effect and the like, meets the propaganda of supporting the work of scientific and technological specializer, and provides more efficient, convenient and fast scientific and technological specializer service based on networking, informatization, digitization and intellectualization for service objects.

Comet-nong-trust APP and wechat applets:

a Hui nong Xin service entrance is built, information technologies such as network communities, video on demand, mobile internet and the like are introduced into agriculture popularization and service in modes such as WeChat public numbers, small programs, mobile APP and the like, and mobile applications such as agriculture science and technology special dispatcher community service, agriculture technology learning, agriculture information pushing and the like are updated and perfected. On the one hand, the masses can upload the scene pictures, the videos and the text descriptions to the science and technology specializer through the system, the mobile phone of the science and technology specializer can remind the experts of how many problems need to be solved, the science and technology specializer carries out detailed technical guidance through the cloud platform, and accurate service of service contents 'picture and text is in the best' and 'point-to-point' is realized. On the other hand, experts can also release technical problems and solutions to be noticed in recent production in the form of videos, images, texts and the like through the cloud platform, and actively push the technical problems and solutions to users paying attention to the industry through the platform.

An access layer: the access layer is a communication service layer between a system front-end system and a system back-end system, nginx is used as reverse proxy service, the presentation layer and the access layer communicate by using https protocol, and the access layer provides an html page and a respAPI interface for the presentation layer;

an application layer: the application layer is built based on a SprigCloud system architecture, platform service logic is realized, cross-platform and cross-database applications are supported, and the application layer service applications are mainly divided into a science and technology special dispatcher registration management and service effect evaluation system, a science and technology special dispatcher data analysis and display system, a coment-rural communication service system, an agricultural technology resource management system and a service cloud platform management system;

applying a support layer: the application support layer realizes the intercommunication between the infrastructure layer and the application layer by utilizing various universal platforms and provides uniform identity authentication for service application; the main units of the application supporting layer are as follows: the system comprises a password application, a workflow engine, a search engine, a message service engine, a file service, unified log management, unified identity authentication and capability service integration;

and (3) a data layer: the data layer is established on hardware and an operating system and is mainly used for storing basic data of the platform and providing data storage support for the whole application, and the data storage of the project is mainly an Oracle database and assisted by a non-relational database, an elastic search and a memory database redis; the main units of the data layer are as follows: the system comprises a scientific and technological special distributor information database, a comet-nong-letter service information database, a public service information database, a basic information database, a data storage and calculation system, a data acquisition and aggregation and data analysis system;

infrastructure layer: the infrastructure layer comprises a system layer and a network layer, is a foundation for supporting the operation of the information system, comprises hardware equipment and a cloud platform, and is a physical environment for bearing the whole information system; the infrastructure layer is provided by a government cloud;

science and technology special distributor supply and demand docking system: the method comprises the steps of extracting topics of personal information of a science and technology specializer and demand content of a service object by using an LDA model, respectively obtaining probability distribution of the topics, measuring topic similarity of the science and technology specializer and user demands through Jensen-Shannon divergence, carrying out demand matching, and carrying out sequencing recommendation according to the similarity.

The system for registration management and service performance evaluation of the scientific and technological speciality comprises the following main components:

the system comprises a workbench, a kort registration, a basic information management, a kort declaration, a work report, a requirement butt joint, a kort management, an audit management, an expert information, an expert studio, a kort wind collecting management, a kort evaluation and assessment, a unit information, a declaration plan management, a kort declaration module management, a work report template management, a data analysis, a service achievement integral calculation and a unit management; establishing a science and technology special distributor registration management and service effect evaluation system, and realizing registration, selection and audit, work report, work log audit, inquiry and statistics, assessment management, performance evaluation and the like of various science and technology special distributors (provincial department special distributor, three-regional talents, city department special distributor and county department special distributor);

the original technology special distributor real-name registration system is upgraded into a technology special distributor management system with unified province, city and county, the technology special distributor is divided into three levels of province level, city level and county level, and the technology special distributor can be independently managed and audited in each county and city. The user system is perfected, and the statistics of the service conditions of the scientific and technological specializer in different levels and different departments can be finally realized by connecting a digital construction result-a legal person library;

and (5) evaluating the performance of the construction science and technology special dispatcher. And calculating the total annual work performance score of the personal science and technology special distributor by adopting a point-counting method according to the scoring category, the scoring item, the scoring standard and the integral value. The evaluation result provides reference basis for the acceptance and exit of the science and technology speciality.

The scientific and technological speciality data analysis and display system mainly comprises:

the method comprises the following steps of performing analysis on a person assigned by a department, performing analysis on information of the department, performing analysis on a department composition, performing analysis on service distribution of the department, performing analysis on poverty-relief county, performing analysis on assistance effect of the department, performing analysis on city department distribution and performing a department map; a technology dispatcher data analysis and display system is constructed, and a typical demonstration is provided for creating digital technology dispatcher management service. The work effect of the scientific and technological specializer in our province is comprehensively displayed through a big data display technology. Big data analysis shows that the service conditions of scientific and technological special dispatchers in cities, counties and regions of the whole province are included; the gradient and the talent team structure of the scientific and technological special local officer talent team in our province; the distribution condition of science and technology specializers of the parties in various regions of our province and the help and success of the science and technology specializers.

The Hui nong Xin service system mainly comprises:

user registration, consultation watching, expert inquiring, agricultural video, requirement docking, remote consultation, expert mobile studio, personal center, APP application setting, content auditing, service data management and agricultural video management; a service-oriented intelligent comet-agriculture-trust service system for service information release and intelligent service of scientific and technological specializer is built, functions of multi-source information fusion, information real-time release, intelligent service and the like of scientific and technological specializer service are achieved through service aggregation, data resource sharing, public support and networked release of service application, a relatively complex technical service scene is contrasted, a video service function is developed, and farmers can call experts through the system to perform online video consultation. The video service mode has the characteristic of strong interactivity, and experts can comprehensively know the actual conditions in the field, so that technical services can be better developed without the limit of time and space.

The agricultural technology resource management system mainly comprises:

agricultural knowledge management, agricultural knowledge auditing, agricultural picture management, agricultural picture auditing, agricultural training video management, agricultural skill training video auditing, identification management and classification management;

according to the actual needs of the dominant industry of advantages on agricultural technologies, an agricultural technology resource management system is established, aiming at the agricultural technology resource database management system of production technologies of various links from varieties, pesticides, fertilizers, plant protection, cultivation to harvesting and processing in the industries such as vegetables, fruit trees, tea, edible fungi, livestock and poultry, aquatic products and the like, the agricultural technology resource management level and the management efficiency are improved.

The service cloud platform management system mainly comprises:

questionnaire management, content management, user management, log management, user feedback, data dictionary management, external service support and security setting;

the construction service cloud platform management system comprises basic functions required by platforms such as questionnaire survey management, content management, log management, user management, authority management, data dictionary management, safety setting and the like, and provides content and management support for portal websites, wisdom agriculture and telecommunications, platform operation and maintenance and the like.

The application support layer mainly comprises the following types:

micro-service application support: service administration, service link tracing (skywalk), service log collection and analysis based on the SpringCloud micro-service framework;

a middleware: the system comprises a workflow engine, a report engine, a search engine and a message service engine;

the business application class supports applications: the method comprises the steps of unified identity authentication, password application, file service, protocol analysis, unified log management and capability service integration;

accessing a third-party interface: the system comprises a government affair data sharing center interface, an Aliyun object storage product interface, an Aliyun video on demand product interface, an Aliyun sensitive word filtering product interface, an Tengcong real-time audio/video interface, a Tengcong short message service interface, a Tengcong cloud map open platform interface and a Gaudi map open platform interface;

the basic support management system is a basic application system of a science and technology dispatcher service cloud platform, is an important component of the platform, and has the functions of unified user management and identity authentication, data management and exchange, information sharing and the like. The application support system is used for carrying out unified and scientific organization and management on all information resources of the platform, providing an integrated access entrance, a basic system support engine, log management and a butt joint external capability service system for the platform, and providing specific information resources and service data meeting the requirements of the platform according to the identity of a user.

The infrastructure layer built-in hardware equipment comprises: server equipment, storage equipment, network equipment and security equipment; the cloud platform is as follows: computing resource pool, storage resource pool, network resource pool, security resource pool, server virtualization, network virtualization, and storage virtualization.

The application layer is built based on a SprigCloud system architecture and supports cross-platform and cross-database application;

the method provides a simple development mode for configuration management, service administration, circuit breakers, intelligent routing, micro-services, control buses, global locks, decision election, distributed sessions and cluster state management operations related to a micro-service architecture.

The application layer builds a built-in SpringCloud access gateway zuul, service communication, a business engine, service registration and discovery Eureka and a configuration center Config based on a SprigCloud system architecture.

The system is connected with an external network platform through an external server interface/capability open service.

Science and technology specializer supply and demand docking

(1) Supply and demand docking service mode

A technology dispatcher expert database is established on the platform, and the detailed technical expertise, the industry and the area where the expert belongs and some detailed information of individuals of each expert are recorded. The enterprise user can upload own technical requirements on the cloud platform, the system can automatically search and match the technical characteristics and technical achievements of experts according to the technical requirements released on the platform by enterprises and cooperative agencies, the experts are recommended to the user, in addition, when the enterprise releases the requirements, the platform pushes the requirements to relevant experts according to requirement keywords, and the experts can select whether to dock or not according to own conditions. On one hand, a technical demand party can consult or apply results by finding an expert, and on the other hand, the expert can also select an enterprise to carry out technical service or result demonstration in a targeted manner. In the same way, the method can better establish a bridge for the 'bidirectional selection' of the selection and the recognition work of the science and technology specializers, so that the platform has the functions of 'intermediary' of the technology and 'intermediary' of the talents, can prevent simple 'Lalang distribution', realizes the functions of 'intermediary' of the technology and 'intermediary' of the talents, and prevents the platform from being separated from the basic reality.

(2) Realization of supply and demand butt joint

The technology specializer needs to be matched with other users who are different from other online social contacts, the matching purpose is to construct a bridge between a service object and expert users, similarity of the service object and the expert users is calculated to find the most suitable two parties by extracting account characteristics of the service object and the expert users, and user matching is completed. All account users of the science and technology specializer service cloud platform are real-name systems, all homepage information is checked manually, authenticity of data can be guaranteed, relevant online social user matching algorithms are compared, a user matching technology based on a fusion multi-attribute algorithm of user file information is suitable for the platform, various user attribute information is taken as character strings to be processed, and an LDA (latent dirichletAllocation) model [16] is used for extracting topics of personal information of the science and technology specializer and required contents of service objects, and probability distribution of the topics of personal information of the science and technology specializer and the required contents of the service objects is obtained respectively. For the descriptions x and y of the science and technology specializer and the user requirements, the topic similarity is measured through Jensen-Shannon divergence, requirement matching is carried out, sequencing recommendation is carried out according to the similarity, and a pairing algorithm of a supply and demand docking system of the science and technology specializer is as follows:

wherein

Is the subject distribution of x and y described by the technology specializer and user requirements,

is the distribution Jensen-Shannon divergence with the value range of [0, 1%]The larger the value is, the higher the degree of dispersion is, and the calculation formula is as follows:

KL (Kullback-Liebler divergence) divergence was:

the network platform mainly comprises: the system comprises a weather station, a government affair information sharing platform, an Ali cloud platform, an Tengcong cloud platform and a rural scientific and technological information resource sharing and service platform.

The whole technical measures are as follows:

the Spring cloud is a micro-service architecture development tool realized based on Spring Boot, and is used for decoupling a complex system and deploying the complex system into independent micro-services. The method provides a simple development mode for the operations of configuration management, service management, circuit breakers, intelligent routing, micro-services, control buses, global locks, decision election, distributed sessions, cluster state management and the like in a micro-service architecture;

spring Boot is a brand-new framework provided by Pivotal team, and the design purpose of the Spring Boot is to simplify the initial building and development process of new Spring application; the framework uses a specific way to configure, thereby eliminating the need for developers to define a templated configuration.

Spring Boot simplifies the application development based on Spring, and an independent Spring application at a product level can be created through a small amount of codes; the Spring Boot provides the setting of opening the box and using the box for the Spring platform and the third-party library, so that the user can start in order. Most Spring Boot applications require only a few Spring configurations.

It uses the concept of "custom over configuration" (there are a lot of configurations in the project, and in addition, a custom configuration is built in to let you not have to configure manually) to let your project run quickly. A project which runs independently (runs jar, is embedded with Servlet container) and is ready to produce and is based on a Spring framework is easy to create by using the Spring Boot, and you can use no Spring or only needs little Spring configuration by using the Spring Boot.

The micro-services developed based on the spring boot are service units which can be independently deployed, horizontally expanded and independently accessed (or have independent databases), the spring cloud is a large manager of the micro-services, after the micro-services are adopted, the number of projects is very large, and the spring cloud as the large manager needs to provide various schemes to maintain the whole ecology.

The advent of SpringCloud can be said to be a huge support and powerful technical background for micro-service architecture. Unlike the general micro-service framework, it only solves a certain problem in the micro-service, but a comprehensive solution framework for the micro-service architecture implementation, which integrates a plurality of widely practiced and proven frameworks as the basic components of the implementation, and creates some very excellent edge components on the basis of the system.

The core function and idea of the SpringBoot are as follows:

spring project run independently: the Spring Boot can be independently operated in the form of a jar packet, and the operation of one Spring Boot project only needs to be performed through java-jar xx.

Embedding Servlet container: spring Boot can choose to embed Tomcat, Jetty, or underrow, so that the project does not have to be deployed in the form of a war package.

Provide starer simplified Maven configuration: spring provides a series of starterpom to simplify the dependent loading of Maven.

Automatic Spring configuration: the Spring Boot automatically configures the beans for the classes in the jar package according to the jar package and the classes in the class path, so that the configuration to be used is greatly reduced. Of course, Spring boots only consider most development scenarios, not all scenarios, and if automatic Bean configuration is needed in actual development and Spring boots do not provide support, automatic configuration can be customized.

The application monitoring Spring Boot available for production provides monitoring of the project at runtime based on http, ssh, telnet.

The method for realizing the Spring Boot without the code generation and the xml configuration is not realized by means of the code generation, but realized by condition annotation, which is a new characteristic provided by Spring 4.x, and the Spring Boot can realize all the configurations of the Spring without any xml configuration.

③ SpringCloud is an ordered set of a series of frames. The development convenience of the Spring Boot is utilized to skillfully simplify the development of infrastructure of a distributed system, such as service discovery registration, configuration center, message bus, load balancing, circuit breaker, data monitoring and the like, and the Spring Boot can be used for one-key starting and deployment in a development style. The Spring does not repeatedly manufacture wheels, only combines more mature service frames which are developed by various companies at present and can withstand practical tests, and encapsulates and shields complex configuration and implementation principles through the Spring Boot style, so that a set of distributed system development toolkit which is simple and easy to understand, deploy and maintain is finally set for developers.

And fourthly, registering and discovering the spring cloud Eureka, wherein the spring cloud Eureka is a part of a spring CloudNetflix micro-service suite, and is subjected to secondary packaging based on the Netflix Eureka and mainly responsible for finishing the service management function in the micro-service architecture. Service governance includes service registration and service discovery.

Service registration: in a service administration framework, a registry is usually constructed, each service unit registers services provided by itself to the registry, the registry is informed of additional information such as a host, a port number, a version number, a communication protocol and the like, and the registry organizes a service list according to service name classification.

Service discovery: since operating under the service administration framework, calls between services are no longer made by specifying specific instance addresses, but rather by making request calls to service names. Therefore, the service invoker does not know the specific service instance location when invoking the service provider interface. Therefore, the caller needs to consult the service registry for services and obtain a list of all the service instances to access a specific service instance.

The Eureka server, also called a service registry. It supports high availability configurations as other service registries do. The method provides good service instance availability by means of strong consistency and can cope with various different fault scenes. If the Eureka is deployed in a cluster mode, when a fragment in the cluster fails, the Eureka goes into a self-protection mode. It allows discovery and registration of services to continue during a shard failure, and other shards in the cluster will synchronize their state back again when the failed shard resumes operation.

The Eureka client, mainly handles registration and discovery of services. The client service is embedded in the code of the client application program through a mode of annotation and parameter configuration, and when the application program runs, the Eureka client registers the service provided by the Eureka client with the registration center and periodically sends heartbeats to update the service lease of the Eureka client. Meanwhile, the system can also inquire the currently registered service information from the server side, cache the currently registered service information to the local and periodically refresh the service state; the operating principle of Eureka is shown in fig. 3.

API gateway spring CloudZuul: the service gateway is an indispensable part of the micro-service architecture. In the process of uniformly providing RESTAPI to an external system through a service gateway, the system has the functions of service routing, load balancing, authority control, verification filtering, combination with a service management framework, a fusing mechanism during request forwarding, service aggregation and the like. Zuul in SpringCloudNeflix plays a role, provides a front door protection function for a micro-service architecture, and simultaneously migrates heavier non-service logic contents controlled by authority to a service routing layer, so that a service cluster main body can have higher reusability and testability.

Spring cloud Zuul registers itself as an application under Eureka service administration by integrating with spring cloud eruuk, while obtaining instance information of all other micro services from Eureka. The design skillfully utilizes the instance information maintained in the service management system, so that the work of maintaining the instance is automatically completed by a service management framework without manual intervention. For the regular maintenance of the route, the Zuul default will create the route mapping by using the service name as the ContextPath, and in most cases, such default setting can already realize most of the routing requirements, and some special configuration is needed except some special cases. But compared with the operation and maintenance workload under the prior architecture, the workload can be greatly reduced after the API gateway is realized by introducing SpringCloudZuul.

The redundancy problem in the micro-service architecture is solved for similar signature check and login check. Theoretically, the check logics are not in much relation with the services of the microservice application per se, so that the check logics can be completely independent into a single service to exist, and only after the check logics are stripped and independent, the check logics are not called for each microservice, but are uniformly called on the API gateway service to perform pre-filtering on the microservice interface so as to realize interception and check on the microservice interface.

SpringCloud Zuul provides a set of filter mechanisms that can support such tasks well. The developer may create various check filters by using Zuul, and then specify which rule requests require the execution of check logic, only those that pass the check will be routed to the particular microservice interface, or else an error prompt is returned. Through the improvement, the micro-service application of each service layer does not need the check logic of non-service property any more, so that the micro-service application can be more concentrated on the development of the service logic, and meanwhile, the automatic test of the micro-service is easier to realize;

sixthly, configuring a center spring cloud Config: the spring cloud Config is a configuration center in the general sense, and the configuration originally placed in the local file by the application is extracted and placed in a central server, so that better management and distribution capability can be provided. The spring cloud Config is divided into a server and a client, wherein the server is also called a distributed configuration center and is an independent micro-service application used for connecting a configuration warehouse and providing access interfaces for the client to acquire configuration information, encryption/decryption information and the like, and the server is responsible for issuing configuration files stored in git (SVN) into REST interfaces. The client is an infrastructure of each service application in the micro-service architecture, manages configuration contents of application resources and services through a specified configuration center, acquires and loads configuration information from the configuration center when the client is started, and can acquire configuration from a REST interface of the server. However, the client cannot actively sense the change of the configuration, and thus actively acquire the new configuration, which requires each client to trigger the respective/refresh through the POST method.

The Spring cloud Config implements abstract mapping of the configuration of the environment variables and attributes in the server and the client, so that it can be used in any application program running in other languages besides the application program built by Spring. Since the configuration center implemented by the SpringCloud Config stores configuration information by default using Git, the configuration server constructed using the SpringCloud Config naturally supports version management of micro-service application configuration information, and configuration contents can be conveniently managed and accessed through a Git client tool. Of course, it also provides support for other storage means, such as SVN repositories, localized file systems; the operation principle of the configuration center is shown in fig. 4.

Seventhly, calling SpringCloudFeign by remote service:

the SpringCloud Feign is an declarative Web Service client, which makes writing Web Service client simpler. This can be done by simply using Feign to create an interface and configuring it with annotations. It has pluggable annotation support, including Feign annotation and JAX-RS annotation. Feign also supports pluggable encoders and decoders. Spring cloud adds support for Feign for Spring MVC annotation, and also integrates Ribbon and Eureka to provide a load balancing HTTP client implementation.

The service load balancing method comprises the following steps:

the SpringCloud Ribbon is a client load balancing tool based on HTTP and TCP, and is realized based on Netflix Ribbon. Through the encapsulation of the SpringCloud, the REST template request facing the service can be easily and automatically converted into the service call with balanced client load.

The Ribbon client component provides a sophisticated set of configuration options such as connection timeouts, retries, retry algorithms, etc. Ribbon embeds pluggable, customizable load balancing components. The following are some load balancing strategies used:

simple polling load balancing;

load balancing of weighted response time;

area-aware polling load balancing;

and (4) random load balancing.

Ninthly, message bus SprincCloudBus:

SpringCloudbus is a message bus component in the SpringCloud architecture that connects various distributed nodes through lightweight message agents. This may be used to broadcast a change in status (e.g., a configuration change) or other message instruction. One core idea of the Spring cloud bus is to extend the Spring Boot application through a distributed initiator, and may also be used to establish a communication channel between multiple applications. Spring CloudBus may use message queue middleware such as AMQP message broker (RabbitMQ), kafka, rocktmq, etc. as a channel.

The essence of spring cloud Buss is to propagate messages in a distributed system using MQ broadcast mechanism, with Kafka and RabbitMQ being commonly used today. By utilizing the bus mechanism, a plurality of things can be done, such as configuration center client real-time refreshing, service event notification, data change event notification and the like.

The application of the spring cloud Bus enables the service application to run on the event trigger model, reduces the observation of the object state actively by developers, and only needs to receive the event notification sent by the changed object, thereby being more concentrated on the realization of the service logic, as shown in fig. 5, the working principle of the application scene of the spring cloud Bus refreshing the new configuration without stopping the application;

in fig. 5, config-repo is the configuration repository, and configServer serves the configuration center; ClineA, ClineB and ClinentC are different running examples of the same application. The steps of using the spring CLOoud Bus for configuration synchronization are as follows:

submitting a code triggering post request, and sending bus/refresh to the ClientA;

the ClientA receives a request to update configuration from the configServer and sends the request to the Spring Cloud Bus;

the Spring Cloudbus receives the message and broadcasts the message to other clients (ClientB, ClientC);

other clients receive the notification and request the configServer to acquire the latest configuration;

all clients acquire the latest configuration.

The r service link trace SkyWalking:

with the popularity of micro-service architecture, some problems under the micro-service architecture are more and more prominent, for example, one request may involve multiple services, and the service itself may depend on other services, the whole request path constitutes a mesh call chain, and once an exception occurs at a certain node in the whole call chain, the stability of the whole call chain is affected.

In view of the above, tools are needed to help understand system behavior and to analyze performance problems so that problems can be quickly located and resolved in the event of a failure.

The current main APM tools comprise Cat, Zipkin, Pinpoint and Skywalk, wherein the Skywalk is used in the project and comprises service management functions such as distributed tracking, performance index analysis, application and service dependence analysis and the like, the zero-intrusion monitoring of an application program is realized through a data acquisition agent component, the analysis granularity is detailed, the support is very rich, and besides the support of conventional Java application monitoring, the monitoring of applications realized by various languages such as Net, node. js, python, php and the like and the monitoring of operation parameters such as a database, a host computer and a docker container are also supported; various backend data storage schemes are supported, such as MySQL, ElasticSearch, H2, and the like.

The core of the Skywalking is a storage platform of data analysis and measurement results, analysis and measurement data are submitted to a Skywalking collector in an HTTP or gRPC mode, the Skywalking collector analyzes and aggregates the data, the data are stored in databases such as elastic search, H2, MySQL, TiDB and the like, and finally the final results can be viewed through a visual interface of a Skywalking UI, and the Skywalking supports the collection of data from multiple sources and multiple formats: skywalk agent in multiple languages, Zipkin v1/v2, Istio surveys, Envoy metrics, and the like.

Front-end system overall technical framework:

platform front-end applications (including PC management platforms, comet-agriculture-trust applets and comet-agriculture-trust APPs) are constructed based on vue.js + ElementUI libraries.

Js is a set of progressive frameworks used to build user interfaces. Js is designed to be applicable layer by layer from bottom to top, unlike other large frameworks. Js' core library only concerns the view layer, is easy to use and is convenient to integrate with a third-party library or an existing project. Js, on the other hand, is also fully capable of providing a driver for complex single-page applications when used in conjunction with modern tool chains and various supporting class libraries.

The big characteristic of vue.js: responsive programming, componentization

Js advantages: lightweight framework, simplicity and easiness in learning, bidirectional data binding, componentization, view, separation of data and structure, virtual DOM and high running speed

Js is single-page application, so that the page is locally refreshed, all data and dom are not required to be requested when the page is skipped every time, the access speed is greatly increased, and the user experience is improved. And his third party ui library saves much development time.

The key technical measures are as follows:

full-text search engine:

the platform application can generate a large amount of valuable data, and the platform can integrate the data and provide a convenient and quick retrieval mode by adopting a mature full-text retrieval product, thereby being beneficial to improving the working efficiency. Full-text retrieval generally includes two elements in form: the method comprises the following steps of searching an entry and a searching result page, but is relatively complex in a background architecture, and the core requirements of the method comprise: chinese word segmentation technology, page capture technology, index establishment, search result sequencing, and statistics, analysis, association, recommendation and the like of search keywords.

Introduction of elastic search:

this project uses the Elasticsearch as a full text search engine. A distributed multi-user capable full-text search engine is provided, based on a RESTful web interface. The Elasticisearch is developed by using Java language and issued as an open source code under Apache licensing terms, is used in cloud computing, can achieve real-time search, and is stable, reliable, rapid, convenient to install and use. It can conveniently make a large amount of data have the capability of searching, analyzing and exploring. The horizontal flexibility of the elastic search is fully utilized, so that the data becomes more valuable in a production environment. The implementation principle of the Elasticissearch is mainly divided into the following steps, firstly, a user submits data to an Elasticissearch database, then a word controller divides words of corresponding sentences, the weights and word division results are stored into the data, when the user searches data, the results are ranked and scored according to the weights, and then returned results are presented to the user.

The method combines the use of Chinese word segmentation plug-ins, has an all-round full-text retrieval function, supports various retrieval modes such as simplicity, combination and expansion and different retrieval methods such as accuracy and fuzziness, provides various retrieval entries such as date, title, keyword and size, and can realize multi-combination multi-level retrieval.

Simultaneously supporting two retrieval modes of a front end and a back end;

supporting both structured data and unstructured data, supporting multiple data types (such as dates, strings, phrases, documents, and binary multimedia types);

supporting the warehousing and retrieval of files with common formats, such as TEXT, HTML, RTF, MSoffice, PDF, S2/PS2/PS, MARC, ISO2709 and the like);

supporting management of multimedia data;

the data storage method supports various forms of data storage modes, such as centralized storage, distributed storage or URL (uniform resource locator) storage, and the data can also be stored in other databases or application systems;

efficient data and index compression, low spatial expansion rate is achieved;

unique (Unique) properties of the support field;

supporting the operation of adding, deleting and modifying the database records;

the library structure is freely copied among different operating system platforms, so that the operation of an administrator is facilitated;

the administrator can set the time of timing optimization, so that the cost of system management is reduced;

when the record is modified, the moving track of the record number is saved, the record can be read by using the record number before moving, and the correctness and the safety of data maintenance are ensured;

various dictionaries can be managed and maintained conveniently;

convenient data backup and recovery functions;

the data export function with various formats can export data type files such as XML and the like.

The full text search engine is applied to the project:

in the project of the service cloud platform of the kort dispatcher, experts matched with keywords are recommended to the user according to the required keywords provided by the user. Aiming at the service scene, the full-text retrieval function of the Elasticissearch is used for realizing the function of the requirement recommendation expert, and the Elasticissearch is adopted because most relational databases do not support full-text retrieval and few relational databases support the low query efficiency of the full-text retrieval, the Elasticisearch can quickly and efficiently carry out full-text retrieval on data by utilizing the characteristic of inverted index, and can expand through simple plug-ins to be matched with ik Chinese participle plug-ins of a third party, chinese word segmentation is carried out on expert data, an expert meeting requirements can be recommended to a user more accurately, a set of relevant scoring standards are realized inside an elastic search, different weight values are configured for different attributes, for example, technical expertise of the expert should be used as the most important basis for recommendation, the weight of technical expertise can be increased, and experts with higher relevance and more meeting the requirements of users are recommended for the users.

The method for recommending the expert for the user requirement by utilizing the Elasticissearch to store the expert data has the advantages that the expert data is firstly synchronized into the Elasticissearch, a timing task synchronization program is compiled by a synchronization scheme according to the business rules to inquire the full amount of the expert data to be synchronized into the Elasticissearch in a timing mode, the synchronous data and the business can be decoupled, normal business is not influenced, the full amount of synchronization mode can reliably ensure the consistency of the data without considering the time sequence problem of data change, the real-time property of the data can also meet the business scene of the expert needing to be recommended, meanwhile, the self-compiled synchronization program is good in flexibility and more convenient for later maintenance, an official logstack data synchronization plug-in can only carry out simple data synchronization and is not flexible enough to meet the synchronization requirement, the incremental synchronization mode is difficult to keep the consistency of the data, and the data synchronization entrance is dispersed in each business, the code coupling is too high, and the later maintenance is difficult, so in conclusion, the expert data is completely synchronized by adopting a self-written synchronization program.

Fourthly, a business process engine:

the science and technology specializer service cloud platform project comprises a large amount of flow business: the method comprises the following steps of expert authentication and audit, unit authentication and audit, subject declaration and audit, work report audit, content compliance audit, agricultural technology resource audit, portal article audit and the like, and a unified flow is needed for control in both daily operation and quick response. Because the auditing process of the platform has high degree of freedom and a plurality of customized processes, a development team realizes a highly customized workflow engine, the configuration and the customization are properly balanced, and the process service of the system is realized. The functions of the device comprise:

the bottom layer provides uniform process data management through data model abstraction, and the process data is separated from the service data, so that the process control engine can adapt to various service scenes;

defining a process;

defining a flow link;

issuing a process;

flow control (checking is passed, not passed, returned, application returned, returned checking and the like);

and inquiring and displaying the process data.

Fifthly, real-time audio and video conversation:

the expert real-time audio and video consultation application is provided in the comet-nong letter citation (including small programs and APPs), the professionality and complexity of a real-time audio and video technology are high, real-time audio and video are integrated through a PaaS service provider, and the App is developed quickly and is a priority choice of a developer at present. Whether the selected RTC is easy to use and fit with the required scene directly influences the project development progress and the later operation and maintenance cost.

Sixthly, product type selection:

basis of product type selection

The real-time audio and video is required to be combined with application message service pushing cloud service at a mobile terminal to achieve real-time audio and video message consultation notification.

Several considerations are taken into account in product model selection:

audio and video frequency talking quality

The three quality indexes of the real-time audio and video are delay, pause and jitter.

The low delay needs to be solved by two aspects, one is a transmission protocol, and the other is to optimize the whole transmission link. The main stream transmission protocols of the real-time audio and video comprise RTMP and UDP, one supports CDN technology and one supports WebRTC technology, relatively speaking, the CDN technology has the time delay of 3-5 seconds, the WebRTC can be within hundreds of milliseconds, and a plurality of manufacturers can simultaneously support the two technologies at present and are respectively suitable for different scenes.

In addition, data packets often arrive in the wrong order, creating jitter related problems, or are lost directly, creating audio-video blanks. Google shows that there are packet loss, excessive jitter or network delay conditions in Duo 99% of all calls of the video chat application. More than 3% of the audio is lost in 20% of calls, and more than 8% of the packet loss rate in 10% of calls, that is, many audio needs to be replaced in each call.

To deal with the above problems, many manufacturers will adopt various mechanisms such as NACK (packet loss retransmission), FEC (forward error correction), adaptive bandwidth adjustment (dynamic adjustment code), receiver Jitter Buffer (media stream stabilization) with packet loss resistance and network Jitter resistance, some of which are used in combination and some of which are used separately, so as to achieve deep understanding before model selection.

Message push quality

The message pushing of the mobile terminal is the basis for establishing audio and video voice communication between the mobile terminal and other terminals, and the delivery rate and stability of the message need to be ensured.

Compatible use platform

Platform compatibility and usability are also important, and differences of all ends are also considered, so that adaptation and characteristic development can be performed on differences of different using ends, and the project development needs to be compatible with multiple applications and multiple equipment manufacturers as shown in the following table:

iOS	Iphone
		Android	mainstream Android manufacturers such as Huashi, millet, charm and the like
WeChat applet	Most of the equipment

Improved development service and document

In some communities, it is common to see some technical documentation that developers ask questions without replying. Developers tend to finish work by themselves more and more in order to improve development efficiency, so that whether development documents are easy to understand and whether Demo is easy to use are very important.

In addition, the work order reply speed, the micro-trust group, the community watching and the response degree and the like can reflect the strength of the PaaS manufacturer service consciousness. Generally, a 7 × 24-hour technical support service, a 1-hour work order fast recovery, a fast remote access, and a fast recovery emergency response mechanism are all very well-established service support for developers.

Results of model selection

By combining the above consideration factors, the project finally selects Tencent Real-Time Communication (TRTC) + as the bottom technical service of Real-Time voice consultation and Real-Time video consultation.

Tencent real-time audio-video product introduction

Tencent Real-Time audio and video (TRTC) can deeply accumulate Tencent in network and audio and video technology for 21 years, and a scheme with two scenes of multi-person audio and video conversation and low-delay interactive live broadcast is opened to developers through Tencent cloud service, so that the Tencent Real-Time audio and video (TRTC) helps the developers to quickly build a low-cost, low-delay and high-quality audio and video interactive solution.

Product architecture

The real-time audio and video TRTC master makes a multi-user audio and video call and low-delay interactive live broadcast solution of full platform intercommunication, and provides SDKs of platforms such as a small program, a desktop browser, Android, iOS, an Electron, Windows, macOS and the like, so that developers can conveniently and quickly integrate and communicate with a real-time audio and video TRTC cloud service background. Through mutual linkage among different products of Tencent cloud, real-time audio and video TRTC can be simply and quickly used in cooperation with cloud products such as instant messaging IM (instant messaging), cloud live broadcast CSS (cascading Style sheets), cloud on-demand VOD (video on demand) and the like, more service scenes are expanded, the product architecture is shown in FIG. 6, and the support platform is shown in the following table:

the product properties are shown in the following table:

the support platform is shown in the following table:

high stability: multiple machine rooms and multiple clusters are deployed, and system stability is guaranteed

High safety: support multiple modes of authentication and encryption

Efficient access: the system is simple and easy to use, and provides 8 language service terminals SDK integrated access such as Java, PHP, C #, Python and the like, and professional technical support team 7x24 hour special line service.

The business process comprises the following steps:

and a message pushing mechanism established by http and websocket and the push is used for transmitting a call state, personal information, a room number and the like required by the audio and video call. The initiator and the receiver use the same room number to join the Tencent cloud audio/video room to establish a video call and voice communication, and fig. 7 shows a general service flow chart of the real-time audio/video call.

And (3) initiating a call:

when a public user initiates an audio and video call consultation request to an expert, the expert needs to be informed in different modes according to the online states of an expert applet and an APP;

when the expert is online:

when an expert is online (a comet-farming small program or a comet-farming APP, hereinafter referred to as a client), the client establishes a websocket connection with the background service system. When the public user initiates a call request to the expert, the background service informs the expert client of the call request through the websock interface, and the expert can immediately answer or refuse the call. The specific business process is shown in fig. 8.

When the expert is offline:

when the expert is offline (the client is not opened), the background can inform the client of the call request only through WeChat service notification and APP push. If the expert logs in the small program and also logs in the APP, the expert is informed of the WeChat service and the APP message at the same time. The expert can select to enter the WeChat small program to answer the call through the WeChat service notice and also can enter the Hui nong Xin APP to answer the call through the APP information. The business flow is shown in fig. 9;

the calling party actively hangs up:

when the calling party hangs up the call, the system needs to judge whether the expert opens the client. If the expert opens the client, the expert is informed to end the call through the websocket service, and the call interface is quitted. If the expert does not open the client, when the expert enters the call through the message, the expert receives the notification that the calling party cancels the call and does not enter the room again, as shown in fig. 10 and 11;

the called party actively hangs up:

when the called party (expert) hangs up actively, the calling party (public) is always on line, so the background service only needs to inform the calling party to end the call through the websocket interface, as shown in fig. 12;

and call recording:

real-time audio and video calls can generate call records in the system. There are 3 nodes for call log data change:

the calling party initiates: and establishing a call record and filling the calling initiation time.

Answering by the called party: and updating the answering time in the call record.

And (4) ending the call: the call record has two forms, one is that the called party does not answer when the time is out, but the called party answers, and hangs up after the consultation is finished. And updating the call ending time in the call record.

End of call time-listen time, i.e. the duration of the call.

And (3) processing an abnormal condition:

if network abnormality occurs in the conversation process, the conversation is abnormally terminated, and the background does not receive a conversation ending request, so that the conversation state of the expert is hung up, at the moment, the expert can not receive the conversation request any more, and the expert needs to log out of the client to log in again and reset the conversation state.

Protecting the privacy number:

the number privacy of experts is not protected, the system provides a telephone number privacy protection function, the public cannot directly see the telephone number of the expert at a comet-nong-letter client (a WeChat applet and an APP), when the public needs to contact the expert in a telephone mode, telephone consultation needs to be selected in an expert studio of the comet-nong-letter client, and the telephone consultation is initiated to the expert through the protection number;

the number privacy protection product of the Aliyun is used for providing privacy communication service. The Ali cloud privacy number protection provides product types such as an AXB intermediate number, an AXG privacy number, an AXN extension number and the like, and through service scene analysis, an AXB intermediate number product is used in the project, as shown in FIG. 13.

Ninthly, binding the privacy number: when a public user initiates a telephone consultation with an expert, the public user needs to acquire a privacy number (AXB intermediate number) first. After receiving a privacy number acquisition request of the front end, the background judges whether the expert has unreleased binding relations, if so, the system prompts that the expert is busy when other people are consulting (in a call) or preparing to consult the expert (the privacy number is acquired and not dialed), otherwise, the system initiates an AXB intermediate number binding request to Ali cloud, establishes a privacy number relation between the dialed number of the public and the expert, and returns the privacy number to the front end. In the present system, the public number is the a number of the AXB, the specialist number is the B number, and the intermediate number is not designated. And randomly acquiring from the number pool. And hiding the expert number and displaying the public number by setting the mode as 3, namely, the private number is dialed by the public user and the real number of the public user is seen by the expert side.

After the number is bound, the background service records the binding relationship and generates an initialization ticket, the call start time and the call end time of the ticket are both null, and the public user is marked not to dial, as shown in fig. 14.

Dialing:

after the public user acquires the privacy number, the comet-nong-letter client automatically wakes up the system to dial the number, and the user can directly dial the privacy number. After dialing, the arilocos will send a dialing notification to the system background service, and the background service records the dialing time, as shown in fig. 15.

And (4) canceling the call:

after the privacy number is acquired, the iOS APP can select to cancel the call, and after the user cancels the call, the system background service automatically releases the binding relationship and updates the binding relationship to be in a failure state, as shown in fig. 16.

And (4) ending the call:

the following cases are all considered as the call end:

the called party is not answering

Refusing to answer by called party

The called party answers and hangs up after the conversation is finished

After the call is finished, the aristoloc sends a call end event to the platform background service, and after the platform receives the notification, the platform updates the call records (answer time, end time and call duration) and initiates a request for releasing the binding relationship to the aristoloc, as shown in fig. 17.

Call log processing

When a calling party acquires a privacy number, a background generates a blank call record, records the time when a user acquires the privacy number, and after the user dials, the Aliyun pushes a dialing notice, and the system updates the dialing time

After the call is finished, the Aliyun pushes a call finish notice, and the system updates the connection time, the call finish time and the call duration.

And (3) abnormal condition processing:

after the user acquires the privacy number, if the user does not cancel the call or dial the call, the system cannot actively release the number binding relationship, and the system needs to automatically release the number binding relationship after the binding relationship fails (the effective time of the binding relationship is set to be 5 minutes at present).

The system queries the number binding relation state from the Aliskiren cloud at regular time (once every 2 minutes), and if the platform is released, the local binding relation table is updated, so that the condition that the expert number is occupied for a long time is avoided.

Although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that changes may be made in the embodiments and/or equivalents thereof without departing from the spirit and scope of the invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (9)

1. The utility model provides a science and technology special dispatcher service cloud platform system of constructing based on big data which characterized in that: the system target architecture is divided into six layers: the system comprises an infrastructure layer, a data layer, an application supporting layer, an application layer, an access layer, a display layer and a science and technology special distributor supply and demand docking system;

a display layer: the display layer is an entrance of the platform facing to the final user, is a main entrance and an interactive interface for various users to obtain required services, and realizes login of user access, integrated access of various business applications, personalized display, user and safety management and other aspects, and is developed based on html5, css, vue.js and G6 languages and a technical framework; the display layer is mainly applied as follows: PC end page, WeChat applet, APP (including Android version and iOS version);

an access layer: the access layer is a communication service layer between a system front-end system and a system back-end system, nginx is used as reverse proxy service, the presentation layer and the access layer communicate by using https protocol, and the access layer provides an html page and a respAPI interface for the presentation layer;

an application layer: the application layer is built based on a SprigCloud system architecture, platform service logic is realized, cross-platform and cross-database applications are supported, and the application layer service applications are mainly divided into a science and technology special dispatcher registration management and service effect evaluation system, a science and technology special dispatcher data analysis and display system, a coment-rural communication service system, an agricultural technology resource management system and a service cloud platform management system;

applying a support layer: the application support layer realizes the intercommunication between the infrastructure layer and the application layer by utilizing various universal platforms and provides uniform identity authentication for service application; the main units of the application supporting layer are as follows: the system comprises a password application, a workflow engine, a search engine, a message service engine, a file service, unified log management, unified identity authentication and capability service integration;

and (3) a data layer: the data layer is established on hardware and an operating system and is mainly used for storing basic data of the platform and providing data storage support for the whole application, and the data storage of the project is mainly an Oracle database and assisted by a non-relational database, an elastic search and a memory database redis; the main units of the data layer are as follows: the system comprises a scientific and technological special distributor information database, a comet-nong-letter service information database, a public service information database, a basic information database, a data storage and calculation system, a data acquisition and aggregation and data analysis system;

infrastructure layer: the infrastructure layer comprises a system layer and a network layer, is a foundation for supporting the operation of the information system, comprises hardware equipment and a cloud platform, and is a physical environment for bearing the whole information system; the infrastructure layer is provided by the government cloud.

Science and technology special distributor supply and demand docking system: the method comprises the steps of extracting topics of personal information of a science and technology specializer and demand content of a service object by using an LDA model, respectively obtaining probability distribution of the topics, measuring topic similarity of the science and technology specializer and user demands through Jensen-Shannon divergence, carrying out demand matching, and carrying out sequencing recommendation according to the similarity.

2. The big data-based science and technology dispatcher service cloud platform construction system as claimed in claim 1, wherein: the system for managing registration of scientific and technological speciality and evaluating service effectiveness mainly comprises the following steps: the system comprises a workbench, a Patty registration, a basic information management, a Patty declaration, a work report, a requirement docking, a Patty management, an audit management, an expert information, an expert studio, a Patty wind collecting management, a Patty evaluation and assessment, a unit information, a declaration plan management, a Patty declaration module management, a work report template management, a data analysis, a service achievement integral calculation and a unit management;

the scientific and technological special dispatcher data analysis and display system mainly comprises: the method comprises the following steps of performing analysis on a person assigned by a department, performing analysis on information of the department, performing analysis on a department composition, performing analysis on service distribution of the department, performing analysis on poverty-relief county, performing analysis on assistance effect of the department, performing analysis on city department distribution and performing a department map;

the comet-nong-letter service system mainly comprises: user registration, consultation watching, expert inquiring, agricultural video, requirement docking, remote consultation, expert mobile studio, personal center, APP application setting, content auditing, service data management and agricultural video management;

the agricultural technology resource management system mainly comprises: agricultural knowledge management, agricultural knowledge auditing, agricultural picture management, agricultural picture auditing, agricultural training video management, agricultural skill training video auditing, identification management and classification management;

the service cloud platform management system mainly comprises: questionnaire management, content management, user management, log management, user feedback, data dictionary management, external service support, and security settings.

3. The big data-based science and technology dispatcher service cloud platform construction system as claimed in claim 2, wherein: the application support layer mainly comprises the following types:

micro-service application support: service administration, service link tracing (skywalk), service log collection and analysis based on the SpringCloud micro-service framework;

a middleware: the system comprises a workflow engine, a report engine, a search engine and a message service engine;

the business application class supports applications: the method comprises the steps of uniform identity authentication, password application, file service, protocol analysis, uniform log management and capability service integration;

accessing a third-party interface: the system comprises a government affair data sharing center interface, an Aliyun object storage product interface, an Aliyun video on demand product interface, an Aliyun sensitive word filtering product interface, an Tencent real-time audio and video interface, an Tencent short message service interface, an Tencent cloud map open platform interface and a Gagde map open platform interface.

4. The big data-based science and technology dispatcher service cloud platform construction system as claimed in claim 1, wherein: the infrastructure layer built-in hardware equipment comprises: the system comprises a server device, a storage device, a network device and a safety device; the cloud platform is as follows: computing resource pool, storage resource pool, network resource pool, security resource pool, server virtualization, network virtualization, and storage virtualization.

5. The big data-based science and technology dispatcher service cloud platform construction system as claimed in claim 1, wherein: the application layer is built based on a SprigCloud system architecture and supports cross-platform and cross-database application;

the method provides a simple development mode for configuration management, service management, circuit breakers, intelligent routing, micro services, control buses, global locks, decision election, distributed sessions and cluster state management operations related to a micro service architecture.

6. The big data-based science and technology dispatcher service cloud platform construction system as claimed in claim 1, wherein: the application layer builds a built-in SpringCloud access gateway zuul, service communication, a service engine, service registration and discovery Eureka and a configuration center Config based on a SprigCloud system architecture.

7. The big data-based science and technology dispatcher service cloud platform construction system as claimed in claim 1, wherein: the system is connected with an external network platform through an external server interface/capability open service.

8. The big data-based science and technology dispatcher service cloud platform construction system as claimed in claim 7, wherein: the network platform mainly comprises: the system comprises a weather station, a government affair information sharing platform, an Ali cloud platform, an Tengcong cloud platform and a rural scientific and technological information resource sharing and service platform.

9. The big data-based science and technology dispatcher service cloud platform construction system as claimed in claim 1, wherein: the pairing algorithm of the science and technology special distributor supply and demand docking system is as follows:

wherein

Is the subject distribution of x and y described by the technology specializer and user requirements,

is the distribution Jensen-Shannon divergence with the value range of [0, 1%]The larger the value is, the higher the degree of dispersion is, and the calculation formula is as follows:

KL (Kullback-Liebler divergence) divergence was:

Images