CN106598585A - Scoring-driven fast service matching and aggregating method in cloud environment - Google Patents

Abstract

The invention discloses a score-driven service rapid matching and aggregation method in a cloud environment, which includes decomposing a target software service into several target service components; searching and matching in a cloud environment to obtain candidate service components for each target service component; Scoring and sorting candidate service components; selecting candidate service components to aggregate to form a preliminary target software service; performing performance testing and correction on the preliminary target software service to obtain the final target software service. The present invention provides that the target software service is disassembled into multiple service components, and the required service components are searched, screened, aggregated and detected in the cloud environment, so as to complete the rapid matching and aggregation of the target service software in the cloud environment, Therefore, the method of the present invention can greatly improve the software service development efficiency in the cloud environment, and at the same time, the method is simple, the feasibility is good, and the reliability is high.

Description

Score-driven fast service matching and aggregation method in cloud environment

technical field

The invention specifically relates to a score-driven service fast matching and aggregation method in a cloud environment.

Background technique

With the development of economy and technology and the increasing popularity of information technology, "cloud" technology has been widely penetrated into people's production and life, bringing endless convenience to people. The so-called "cloud" technology refers to a hosting technology that unifies a series of resources such as hardware, software, and network in a wide area network or a local area network to realize data calculation, storage, processing, and sharing. In the cloud environment, users can obtain massive resources and massive services.

Similarly, with the development of economy and technology, the concept of "tailor-made" has gradually become popular, especially for the software industry with a high degree of personalization and large functional differentiation, "tailor-made" differentiated software With its obvious advantages such as interface customization and function customization, it has been favored by the majority of users.

However, the software industry has ushered in the "tailor-made" era, and it has also ushered in huge problems. The popularity of customized software means that the acceptance of general-purpose software is relatively low, and it also means that the software development cycle is extended: because developers need to redesign the software architecture, software services, and software data for each software. etc., which significantly prolongs the software development cycle and greatly affects the software development efficiency.

Contents of the invention

The purpose of the present invention is to provide a score-driven rapid service matching and aggregation method in the cloud environment, which can greatly improve the development efficiency of software services, and is simple and feasible.

The score-driven service fast matching and aggregation method under the cloud environment provided by the present invention comprises the following steps:

S1. According to the characteristics of the target software service, decompose the target software service into N target service components, and determine the parameters and requirements of each target service component;

S2. Based on cloud computing, search and match N target service components required in step S1 in the cloud environment, and obtain M candidate service components for each target service component;

S3. Scoring the M candidate service components of each target service component obtained in step S2;

S4. According to the scoring result obtained in step S3, sort the M candidate service components of each target service component;

S5. According to the sorting result of the candidate service components of each target service component, select a candidate service component among the candidate service components of each target service component, and aggregate all the selected candidate service components to form a preliminary target software service;

S6. Perform performance detection and correction on the preliminary target software service obtained in step S5, so as to obtain the final target software service.

The parameters and requirements of each target service component described in step S1 specifically include the functional characteristics of the target service component, as well as the type, number, length and precision of the input data of the target service component, and the type, number, length and precision of the output data .

Searching and matching the target service component described in step S2, specifically, the type, number, length and precision of the input data of the search and match in the cloud environment and the target service component, as well as the type, number, length and precision of the output data are the same , and the candidate service components whose functional characteristics are similar to the target service components.

Scoring candidate service components in step S3 is to score candidate service components using fuzzy evaluation rules.

The use of fuzzy evaluation rules to score candidate service components specifically includes the following steps:

1) Select the evaluation index of the candidate service component, said index includes one-class index R=[r ₁ ,r ₂ ... r _n ], and select the second-class index r _i =[r _ij ] for each one-class index, so Said 1≤i≤n;

2) For each second-class index, use the expert system to score, so as to obtain the score S _rij of each second-class index;

3) For each first-class index, set the weight value k _j of the second-class index under the first-class index, so as to obtain the score of each first-class index

4) For each candidate service component, set the weight value q _i of each category index, and calculate the final score of each candidate service component The higher the score, the better the performance of the candidate service component.

The step S5 of selecting a candidate service component among the candidate service components of each target service component is specifically selecting a component with the highest score among the candidate service components of each target service component as the candidate service component.

As described in step S5, the selected candidate service components are aggregated to form a preliminary target software service, which specifically includes the following steps:

A. Adaptation of service components: select the correct parameters to adapt or modify the service components, so that the service components can be applied to software services;

B. Aggregation of service components: On the basis of the service component model, through the service component framework, architecture description language, glue code, scripting language and collaborative language technology, the adapted construction is aggregated into a complete software service.

The performance detection of the target software service described in step S6 specifically includes the following steps:

Ⅰ. Perform performance testing on the target software service;

Ⅱ. Judging the results of performance testing:

If the detection is passed, the target software service aggregation is completed;

If the detection fails, then determine the type of error in the detection;

Ⅲ. Types of judgment errors:

If there is a test error for a certain service component, replace the wrong service component with the service component with the second highest score in step S4;

If an error occurs in the overall test of the software service, among all the service components constituting the software service, the service component with the lowest score is replaced with the service component with the second highest score in step S4;

Ⅳ. Re-aggregate all service components, and perform performance testing on the re-aggregated target software services;

Ⅴ. Repeat steps Ⅰ to Ⅳ until the aggregated target software service passes the performance test; or, if all the service components constituting the target software service have been replaced, it indicates that the rapid matching and aggregation of the service components failed this time, and an alarm is issued, requesting human intervention.

The performance detection of the target software service described in step S6 is specifically the detection of the target software service by using dynamic testing technology and black box testing technology.

The score-driven service rapid matching and aggregation method in the cloud environment provided by the present invention disassembles the target software service into multiple service components, and searches, screens, and aggregates the required service components in the cloud environment and detection, so as to complete the rapid matching and aggregation of target service software in the cloud environment, so the method of the present invention can greatly improve the software service development efficiency in the cloud environment, and at the same time, the method is simple, the feasibility is good, and the reliability is high.

Description of drawings

Fig. 1 is a schematic flow chart of the method of the present invention.

Fig. 2 is a detailed flow diagram of the method of the present invention.

detailed description

As shown in Figure 1, it is a schematic flow diagram of the method of the present invention, and Figure 2 is a schematic flow diagram of the detailed flow of the method of the present invention: the score-driven service fast matching and aggregation method under the cloud environment provided by the present invention includes the following steps:

S1. According to the characteristics of the target software service, decompose the target software service into N target service components, and determine the parameters and requirements of each target service component;

The parameters and requirements of each target service component, specifically including the functional characteristics of the target service component, as well as the type, quantity, length and precision of the input data of the target service component, and the type, quantity, length and precision of the output data, etc.;

S2. Based on cloud computing, search and match N target service components required in step S1 in the cloud environment, and obtain M candidate service components for each target service component;

When searching for a match specifically, the type, number, length and precision of the input data of the search match and the target service component in the cloud environment, as well as the type, number, length and precision of the output data are the same, and the functional characteristics and the target service Candidate service components with similar components;

S3. Scoring the M candidate service components of each target service component obtained in step S2;

Specifically, fuzzy evaluation rules can be used to score candidate service components, which specifically include the following steps:

1) Select the evaluation index of the candidate service component, said index includes one-class index R=[r ₁ ,r ₂ ... r _n ], and select the second-class index r _i =[r _ij ] for each one-class index, so Said 1≤i≤n;

2) For each second-class index, use the expert system to score, so as to obtain the score S _rij of each second-class index;

3) For each first-class index, set the weight value k _j of the second-class index under the first-class index, so as to obtain the score of each first-class index

4) For each candidate service component, set the weight value q _i of each category index, and calculate the final score of each candidate service component The higher the score, the better the performance of the candidate service component.

For example, use an expert system to score each candidate service component according to the indicators described in Table 1 below;

Table 1 Schematic diagram of scoring indicators for candidate service components

Then, for each first-class index, set the weight value of the second-class index under the first-class index, and calculate the score of each first-class index; for a candidate service component, its functional score, the completeness is excellent, That is, 5 points; the score of interoperability is excellent, that is, 5 points; the score of standardization is good, that is, 4 points; then the weight value of the second category index of the functional score of the component is 0.4, and the interoperability 0.2 and standardization 0.4, then the functional score of the component is 5*0.4+5*0.2+4*0.4=4.6; for each candidate service component, set the weight value of each category index, And calculate the final score of each candidate service component; the higher the score, the better the performance of the candidate service component;

S4. According to the scoring result obtained in step S3, sort the M candidate service components of each target service component;

S5. According to the sorting results of the candidate service components of each target service component, select a candidate service component with the highest score among the candidate service components of each target service component, and aggregate all the selected candidate service components to form a preliminary target software service;

When aggregated, it includes two aspects:

A. Adaptation of service components: select the correct parameters to adapt or modify the service components, so that the service components can be applied to software services;

B. Aggregation of service components: On the basis of the service component model, through the service component framework, architecture description language, glue code, scripting language and collaborative language technology, the adapted construction is aggregated into a complete software service;

S6. Perform performance detection and correction on the preliminary target software service obtained in step S5, so as to obtain the final target software service.

When performing performance testing on the target software service, the following steps can be taken specifically:

Ⅰ. Perform performance testing on the target software service;

Ⅱ. Judging the results of performance testing:

If the detection is passed, the target software service aggregation is completed;

If the detection fails, then determine the type of error in the detection;

Ⅲ. Types of judgment errors:

If there is a test error for a certain service component, replace the wrong service component with the service component with the second highest score in step S4;

If an error occurs in the overall test of the software service, among all the service components constituting the software service, the service component with the lowest score is replaced with the service component with the second highest score in step S4;

Ⅳ. Re-aggregate all service components, and perform performance testing on the re-aggregated target software services;

Ⅴ. Repeat steps Ⅰ to Ⅳ until the aggregated target software service passes the performance test; or, if all the service components constituting the target software service have been replaced, it indicates that the rapid matching and aggregation of the service components failed this time, and an alarm is issued, requesting human intervention.

For performance testing of target software services, dynamic testing techniques and black-box testing techniques can be used to test target software services.

The invention patent is supported by the National Natural Science Foundation of China (Project No. 61304184 and Project No. 61672221).

Claims (9)

1. score under a kind of cloud environment the service Rapid matching and polymerization of driving, comprise the steps：

S1. it is N number of destination service component by target software service decomposition according to the characteristic of target software service, and determines each The parameter of destination service component and requirement；

S2. based on cloud computing, the N number of destination service component searched for required for matching step S1 under cloud environment obtains every M candidate service component of individual destination service component；

M candidate service component of each the destination service component for S3. obtaining to step S2 scores；

S4. the appraisal result for being obtained according to step S3, is ranked up to M candidate service component of each destination service component；

S5. according to the ranking results of the candidate service component of each destination service component, in the candidate of each destination service component A candidate service component is chosen in services component, and the candidate service component of all selections is carried out being polymerized forming preliminary mesh Mark software service；

S6. the preliminary target software service for obtaining for step S5 carries out performance detection and amendment, so as to obtain final mesh Mark software service.

2. score under cloud environment according to claim 1 the service Rapid matching and polymerization of driving, it is characterised in that The parameter of each destination service component described in step S1 and requirement, specifically include the functional characteristic of destination service component, and The type of the input data of destination service component, number, length and precision, and the type of output data, number, length and essence Degree.

3. score under cloud environment according to claim 1 the service Rapid matching and polymerization of driving, it is characterised in that Matching is scanned for destination service component described in step S2, search matching and destination service component specially under cloud environment The type of input data, number, length and precision, and the type of output data, number, length and precision all same, and The functional characteristic candidate service component similar with destination service component.

4. score under the cloud environment according to one of claims 1 to 3 the service Rapid matching and polymerization of driving, it is special Levy be scoring that candidate service component is carried out described in step S3 be that candidate service component is carried out using fuzzy evaluation rule Scoring.

5. score under the cloud environment stated according to claim 4 the service Rapid matching and polymerization of driving, it is characterised in that institute The employing fuzzy evaluation rule stated scores candidate service component, specifically includes following steps：

1) evaluation index of candidate service component is chosen, the index includes a class index R=[r₁,r₂…r_n], and to each Class index r of one class selecting index two_i=[r_ij], the 1≤i≤n；

2) for each two class index, scored using specialist system, so as to obtain the scoring of each two class index S_rij；

3) for each class index, the weighted value k of two class indexs under the class index is set_j, so as to obtain each class The score of index

4) each candidate service component is directed to again, set the weighted value q of each class index_i, and calculate each candidate's clothes The final score of business componentThe score is higher, then show that the performance of candidate service component is better.

6. score under cloud environment according to claim 5 the service Rapid matching and polymerization of driving, it is characterised in that A candidate service component is chosen in the candidate service component of each destination service component described in step S5, specially every The component of a highest scoring is chosen in the candidate service component of individual destination service component as candidate service component.

7. score under the cloud environment according to one of claims 1 to 3 the service Rapid matching and polymerization of driving, it is special It is to carry out being polymerized forming preliminary target software service by the candidate service component chosen described in step S5 to levy, and is specifically then included Following steps：

A. the adaptation of services component：Correct parameter is selected to be adapted to services component or change, so that services component can Suitable for software service；

B. the polymerization of services component：On the basis of service-oriented component model, language is described by services component framework, architecture Speech, glue code, script and collaboration language technology, by the structure being adapted to a complete software service is aggregated into.

8. score under the cloud environment according to one of claims 1 to 3 the service Rapid matching and polymerization of driving, it is special It is to carry out performance detection to target software service described in step S6 to levy, and specifically includes following steps：

I. performance detection is carried out to target software service；

II. judge the result of performance detection：

If detection passes through, target software service aggregating is completed；

If detection does not pass through, judge occur the type of mistake in detection；

III. wrongheaded type：

If some services component occurs in that test errors, then the services component that will appear from mistake replaces with score in step S4 Secondary high services component；

If software service integrated testability occurs in that mistake, then by all services components for constituting the software service, score Minimum services component is replaced with and obtain in step S4 services component high by several times；

IV. all of services component is polymerized again, and again to regrouping after target software service carry out performance Detection；

V. I～step IV of repeat step, until the target software service of polymerization passes through performance detection；Or, if it is soft to constitute target The services component of part service is all replaced, then show this services component Rapid matching and the failure that is polymerized, and sends warning, Request manual intervention.

9. score under the cloud environment according to one of claims 1 to 3 the service Rapid matching and polymerization of driving, it is special It is to carry out performance detection to target software service described in step S6 to levy, specially using technique of dynamic measurement and Black-box Testing Technology is detected to target software service.