CN117336352A - Qualitative and quantitative mixed cloud service quality assessment method and system - Google Patents
Qualitative and quantitative mixed cloud service quality assessment method and system Download PDFInfo
- Publication number
- CN117336352A CN117336352A CN202311263567.3A CN202311263567A CN117336352A CN 117336352 A CN117336352 A CN 117336352A CN 202311263567 A CN202311263567 A CN 202311263567A CN 117336352 A CN117336352 A CN 117336352A
- Authority
- CN
- China
- Prior art keywords
- service
- qualitative
- arc
- preference
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001303 quality assessment method Methods 0.000 title claims abstract description 15
- 238000004364 calculation method Methods 0.000 claims abstract description 38
- 238000012360 testing method Methods 0.000 claims abstract description 16
- 238000011158 quantitative evaluation Methods 0.000 claims abstract description 9
- 238000000034 method Methods 0.000 claims description 39
- 239000011159 matrix material Substances 0.000 claims description 21
- 238000011156 evaluation Methods 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 8
- 238000007689 inspection Methods 0.000 claims description 3
- 238000007726 management method Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 7
- 230000002441 reversible effect Effects 0.000 description 6
- 230000006698 induction Effects 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 4
- 238000002474 experimental method Methods 0.000 description 4
- 238000012986 modification Methods 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 238000013441 quality evaluation Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 3
- 238000011144 upstream manufacturing Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 2
- 239000013256 coordination polymer Substances 0.000 description 2
- 238000009795 derivation Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000003672 processing method Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000005251 gamma ray Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/51—Discovery or management thereof, e.g. service location protocol [SLP] or web services
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/50—Testing arrangements
- H04L43/55—Testing of service level quality, e.g. simulating service usage
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a qualitative and quantitative mixed cloud service quality assessment method and a qualitative and quantitative mixed cloud service quality assessment system, wherein the qualitative and quantitative mixed cloud service quality assessment method comprises the following steps: acquiring qualitative service attributes and preference relations of cloud services; consistency test is carried out on the preference relation of the qualitative service attribute to obtain a preference relation after consistency test; modeling according to the qualitative service attribute and the preference relation after consistency test to obtain a graphical model; based on the graphical model, combining the values of different qualitative service attributes to obtain a specific service, and quantitatively evaluating the specific service; according to the quantitative evaluation results of all specific services, calculating to obtain utility values of qualitative service attributes; and carrying out weighted summation on the utility value of the qualitative service attribute and the utility value of the known quantitative service attribute to obtain a comprehensive utility value, and finding out the optimal specific service according to the comprehensive utility value. The invention solves the problem of low calculation efficiency of qualitative service attribute and preference thereof.
Description
Technical Field
The invention relates to the technical field of cloud service quality evaluation, in particular to a qualitative and quantitative mixed cloud service quality evaluation method and system.
Background
With the rapid development of cloud computing, the realization of business process optimization by using cloud services is a necessary trend of enterprise sustainable development and improvement of core competitiveness. Because of the existence of a large number of cloud services with similar functions and different service qualities, how to recommend optimal services to users under the conditions of a rapid increase in the number of services and complex execution environments has become a hot spot of current academic and industrial concern.
Currently, service providers often employ quality of service to describe or measure the performance of a service. According to the quality type, the quality of service can be classified into quantitative quality of service and qualitative quality of service. The quantitative service quality can be expressed in numerical forms such as integer, decimal, percentage, etc., and the common quantitative service attributes such as response time and cost. Current service preference research based on quality of service awareness is mostly focused on quantitative service attributes, which aim to recommend services to users that maximize quality of service performance with meeting constraints. However, in an actual business environment, a part of the quality of service cannot be described in a quantitative form, such as a service storage type, a service deployment manner, and the like. Such attributes are collectively referred to as qualitative service attributes and are generally described in terms of classification values. Rather than quantitative service attributes being explicitly computable, the evaluation of qualitative service quality is typically dependent on user preferences. The user may not only have a preference for the value of a certain qualitative service attribute, but may also have a emphasis on different qualitative service attributes. For example, in terms of service storage type, a user prefers a file system over a database system, and it considers the storage type of a service to be more important than its deployment.
Since quantitative processing models are not suitable for evaluation of qualitative service attributes, an appropriate model representation must be selected and qualitative service quality evaluated. Although some researches propose qualitative preference calculation methods based on CP-nets and TCP-nets, the existing methods are low in efficiency, and the qualitative service attribute performance of acquiring services by adopting a derivation mode based on an induction diagram is not suitable for the situations of more preference relations, excessive attribute values and large data volume. In addition, none of the above studies considers that manually set preference information is extremely likely to cause a situation of preference conflict, redundancy, or the like. Once the wrong preference network is generated, no accurate assessment of the service can be made. Therefore, how to correctly process preference information of a user and implement quality of service evaluation according to the preference information is a key problem to be solved urgently for service preference.
There are few processes in the prior art that consider the qualitative service attributes of a service and their preferences. On the one hand, no consistency check is performed on the qualitative preference relationship proposed by the user, which may lead to a false representation of the user preference information, ultimately leading to quality of service assessment conflict problems. On the other hand, in qualitative service attributes and preference calculation thereof, the existing solution method based on the induction diagram is low in efficiency and poor in expandability, and is difficult to be applied to complex cloud service scenes with complex preference relations, multiple attribute values and large service quantity; the existing method based on the penalty value greatly simplifies quantitative evaluation of qualitative service quality by using the penalty value calculation, but repeated iterative calculation is needed according to a preference relation diagram when calculating the weight of each attribute, so that the algorithm efficiency needs to be further improved.
Disclosure of Invention
Therefore, the technical problem to be solved by the invention is to solve the problem of low efficiency of a solving method in qualitative service attribute and preference calculation thereof in the prior art.
In order to solve the technical problems, the invention provides a qualitative and quantitative mixed cloud service quality evaluation method, which comprises the following steps:
step S1: acquiring qualitative service attributes and preference relations of cloud services;
step S2: consistency test is carried out on the preference relation of the qualitative service attribute to obtain a preference relation after consistency test;
step S3: modeling according to the qualitative service attribute and the preference relation after consistency test to obtain a graphical model;
step S4: based on the graphical model, combining the values of different qualitative service attributes to obtain a specific service, and quantitatively evaluating the specific service;
step S5: according to the quantitative evaluation results of all specific services, calculating to obtain utility values of qualitative service attributes;
step S6: and carrying out weighted summation on the utility value of the qualitative service attribute and the utility value of the known quantitative service attribute to obtain a comprehensive utility value, and finding out the optimal specific service according to the comprehensive utility value.
In one embodiment of the present invention, the preference relationship in the step S1 includes preference of qualitative service attribute value, where the preference of qualitative service attribute value is specifically divided into:
given a qualitative service attribute X i The value range isWherein (1)>Representing qualitative service attributes X i N of (2) i The number is further divided into two types according to whether preference conditions exist, including:
if no preference condition exists, unconditionally prefers UP k The formula is: wherein, > represents the partial order of attribute values;
if a preference condition exists, the CP is preferred for the condition k The formula is: wherein [ Pa (X i ):y]The conditional attribute indicating the preference information is Pa (X i ) The conditional attribute takes the value y.
In one embodiment of the present invention, the preference relationship in the step S1 includes preference among qualitative service attributes, and the preference among qualitative service attributes is specifically divided into:
given two qualitative service attributes X i And X j Further, two categories are classified according to whether or not there is a preference condition, including:
if no preference condition exists, the relative importance RI k The formula is:wherein (1)>Representing a partial order between attributes;
if there is a preference condition, then it is the conditional relative importance CI k The formula is:wherein [ X ] c :y]The conditional attribute representing the relative importance is X c The conditional attribute takes the value y.
In one embodiment of the present invention, in the step S3, modeling is performed according to the qualitative service attribute and the preference relationship after the consistency test, so as to obtain a graphical model, and the method includes:
modeling qualitative service attributes and preference relations after consistency test through a graphical tool TCP-nets to obtain a graphical model G, wherein the graphical model G is expressed as:
G＝<V,cp,ri,ci,cpt,cit>
wherein v= { X 1 ,X 2 ,…,X n A set of nodes, corresponding to a set of qualitative service attributes; cp is a set of conditional preference arcs cp-arc, cp-arc=<X i ,X j >X represents i Is X j And X is the conditional node of (2) i ＝Pa(X j ) The method comprises the steps of carrying out a first treatment on the surface of the ri is a set of relative importance preference arcs ri-arc, ri-arc=<X i ,X j >X represents i Ratio X j More importantly, i.eci is a set of conditional relative importance arcs ci-arc, ci-arc= (X) i ,X j ) X represents i And X j The importance of (2) depends on the value of the condition node; CPT is a set of conditional preference tables CPT, CPT (X i ) For storing X under conditional nodes i A partial order relation of values; CIT is a set of conditional importance tables CIT, CIT (X i ,X j ) For storing X under conditional nodes i And X j Is a partial order relationship of (a).
In one embodiment of the present invention, the step S2 performs consistency check on the preference relationship of the qualitative service attribute, and the method includes:
let gamma 1 And gamma 2 Representing qualitative service attributes X 1 And X 2 Two preference relations, gamma 1 ⊙γ 2 Representing the processing result of the preference relation after consistency check, wherein the result of the consistency check is divided into two types of compatibility and conflict;
the compatibilizing comprises:
if gamma is 1 Is cp-arc:<X 1 ,X 2 >，γ 2 for ri-arc:<X 1 ,X 2 >gamma is then 1 ⊙γ 2 The result of (2) is gamma 1 ；
If gamma is 1 Is cp-arc:<X 2 ,X 1 >，γ 2 for ri-arc:<X 2 ,X 1 >gamma is then 1 ⊙γ 2 The result of (2) is gamma 1 ；
If gamma is 1 Is cp-arc:<X 1 ,X 2 >，γ 2 is ci-arc (X) 1 ,X 2 ) Gamma is then 1 ⊙γ 2 The result of (2) is gamma 1 ；
If gamma is 1 Is cp-arc:<X 2 ,X 1 >，γ 2 is ci-arc (X) 1 ,X 2 ) Gamma is then 1 ⊙γ 2 The result of (2) is gamma 1 ；
The conflict includes:
if gamma is 1 Is cp-arc:<X 1 ,X 2 >，γ 2 is cp-arc:<X 2 ,X 1 >gamma is then 1 ⊙γ 2 The result of (2) is None;
if gamma is 1 For ri-arc:<X 1 ,X 2 >，γ 2 for ri-arc:<X 2 ,X 1 >gamma is then 1 ⊙γ 2 The result of (2) is None;
if gamma is 1 Is cp-arc:<X 1 ,X 2 >，γ 2 for ri-arc:<X 2 ,X 1 >gamma is then 1 ⊙γ 2 The result of (2) is None;
if gamma is 1 For ri-arc:<X 1 ,X 2 >，γ 2 is cp-arc:<X 2 ,X 1 >gamma is then 1 ⊙γ 2 The result of (2) is None;
wherein None represents the need to be to node X 1 And X 2 And the preference relation between the two is described again until the preference relation is valid.
In one embodiment of the present invention, the method combines the values of different qualitative service attributes based on the graphical model to obtain a specific service, and quantitatively evaluates the specific service, where the method includes:
introducing a directed arc sci-arc into the graphical model G:<X i ,X j >indicating the pointing relation between the conditional nodes and the corresponding nodes in the ci-arc to generate a dependency graph G which is used for indicating the topological ordering of all the nodes in the graphical model G;
according to the inverse topological order of the dependency graph G, a node weight list W and an arc value matrix Val are obtained, wherein the arc value matrix Val is used for storing quantized values of each preference relation, and the node weight list W comprises weight values W (X i )；
For a particular service s, calculate a qualitative service attribute X i Rank of values of (2) s (X i ) And simplifying qualitative service attribute X by using node weight list W and arc value matrix Val i Importance weight w of (2) s (X i ) To derive a penalty value pen(s), wherein the specificThe service s is a service obtained by combining different values in different qualitative service attributes, and the qualitative service attributes X i Rank of values of (2) s (X i ) The definition is as follows: suppose qualitative service attribute X i The value range of (2) isRank is then s (X i ) The value range of (2) is [0, |D (X) i )|-1]。
In one embodiment of the present invention, the formula of the arc value matrix Val is:
Val＝[v ij ] n×n
wherein n is the number of qualitative service attributes, v ij Is node X i And X j The value of the arc in between; given two nodes X i And X j And an arc gamma between nodes, v if gamma is any one of directional arcs cp-arc, ri-arc, sci-arc ij Storage Slave X i Pointing to X j And v is the arc value of ji =0; if gamma is the undirected arc ci-arc, then v ij And v ji Separately storeAnd->Arc value in case;
the weight value W (X i ) The formula is:
wherein ri (X) i )，ci(X i )，cp(X i ) And sci (X) i ) From X i Starting ri-arc, ci-arc, cp-arc and sci-arc sets;
v ij ＝w(X j )·|D(X j )|
wherein, |D (X) j ) I is X j Is used to determine the value range size of the (c) in the (c),w(X j ) Is X j Is a node weight of (c).
In one embodiment of the present invention, the penalty value pen(s) is formulated as:
where V is the qualitative service attribute set of a particular service s, rank s (X i ) Representing qualitative service attributes X in a particular service s i Value ordering of w s (X i ) Representing qualitative service attributes X in a particular service s i The importance weights and formulas of (2) are:
wherein ri (X) i )，ci(X i )，cp(X i ) And sci (X) i ) From X i Starting ri-arc, ci-arc, cp-arc and sci-arc sets,representing attribute X i Specific attribute X j More important.
In one embodiment of the present invention, in the step S5, according to the quantized evaluation results of all the specific services, a utility value of the qualitative service attribute is calculated, where the formula is:
where pen(s) represents the penalty value of a particular service s, pen max Hepen min Is the largest penalty value and the smallest penalty value in all specific services.
In order to solve the technical problems, the invention provides a qualitative and quantitative mixed cloud service quality evaluation system, which comprises:
the acquisition module is used for: the cloud service preference relationship management method comprises the steps of acquiring qualitative service attributes of cloud services and preference relationships of the qualitative service attributes;
and (3) a checking module: the preference relation is used for carrying out consistency check on the preference relation of the qualitative service attribute to obtain a preference relation after consistency check;
modeling module: the method comprises the steps of modeling according to qualitative service attributes and preference relations after consistency inspection to obtain a graphical model;
and an evaluation module: the method comprises the steps of combining values of different qualitative service attributes based on the graphical model to obtain a specific service, and quantitatively evaluating the specific service;
the calculation module: the utility value of the qualitative service attribute is obtained through calculation according to the quantitative evaluation results of all the specific services;
and (3) optimizing module: and the utility value of the qualitative service attribute and the utility value of the known quantitative service attribute are weighted and summed to obtain a comprehensive utility value, and the optimal specific service is found out according to the comprehensive utility value.
Compared with the prior art, the technical scheme of the invention has the following advantages:
according to the method, consistency check is carried out on the preference relation of the qualitative service attributes in the cloud service, so that the problem of service performance evaluation conflict can be effectively avoided, the calculation flow can be simplified, and the calculation time of qualitative attribute combination preference is saved;
the invention improves the punishment value, solves the problem of repeatedly calculating the preference relation in the traditional method, and facilitates the quantitative calculation of the qualitative service attribute combination of the subsequent specific service.
Drawings
In order that the invention may be more readily understood, a more particular description of the invention will be rendered by reference to specific embodiments thereof that are illustrated in the appended drawings.
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is an exemplary diagram of TCP-nets in an embodiment of the invention;
FIG. 3 is a dependency graph of an exemplary diagram of TCP-nets in accordance with an embodiment of the present invention;
FIG. 4 is a flow chart of calculating penalty values for a particular service according to an embodiment of the invention;
FIG. 5 is a schematic diagram of calculation time of two qualitative attribute solving methods under the condition of changing the number of abstract tasks in the embodiment of the invention;
fig. 6 is a schematic calculation time diagram of two qualitative attribute solving methods in the case of changing the number of candidate services according to the embodiment of the present invention.
Detailed Description
The present invention will be further described with reference to the accompanying drawings and specific examples, which are not intended to be limiting, so that those skilled in the art will better understand the invention and practice it.
Example 1
Referring to fig. 1, the present invention relates to a qualitative and quantitative mixed cloud service quality assessment method, which comprises:
step S1: acquiring qualitative service attributes and preference relations of cloud services;
step S2: consistency test is carried out on the preference relation of the qualitative service attribute to obtain a preference relation after consistency test;
step S3: modeling according to the qualitative service attribute and the preference relation after consistency test to obtain a graphical model;
step S4: based on the graphical model, combining the values of different qualitative service attributes to obtain a specific service, and quantitatively evaluating the specific service;
step S5: according to the quantitative evaluation results of all specific services, calculating to obtain utility values of qualitative service attributes;
step S6: and carrying out weighted summation on the utility value of the qualitative service attribute and the utility value of the known quantitative service attribute to obtain a comprehensive utility value, and finding out the optimal specific service according to the comprehensive utility value.
The present embodiment is described in detail below:
1. formalized problem model
In an actual business environment, the preference relationship of qualitative service attributes can be divided into two categories: the preference to value the service attribute and the preference between the service attributes. Each class may be further subdivided according to whether or not there are preference conditions, which are defined as follows:
definition 1 (preference for service attribute values): given a qualitative service attribute X i The value range isFurther subdivided into two categories depending on whether preference conditions exist: (1) Unconditional preference UP k As shown in formula (1). Wherein > represents the partial order of attribute values; (2) Conditional preference CP k As shown in formula (2). Wherein [ Pa (X) i ):y]The conditional attribute indicating the preference information is Pa (X i ) The conditional attribute takes the value y.
Definition 2 (preference between service attributes): given two qualitative service attributes X i And X j Further subdivided into two categories depending on whether preference conditions exist: (1) Relative importance RI k As shown in formula (3). Wherein the method comprises the steps ofRepresenting a partial order between attributes; (2) Condition relative importance CI k As shown in equation (4). Wherein [ X ] c :y]The conditional attribute indicating the relative importance is X c The conditional attribute takes the value y.
The present embodiment uses TCP-nets to model qualitative service attributes and their preference relationships. TCP-nets is a graphical model whose nodes represent qualitative service attributes and arcs represent various types of preference relationships. In addition, each node corresponds to a Conditional Preference Table (CPT) for representing the preference relationship of the node values, and each undirected arc corresponds to a Conditional Importance Table (CIT) for representing the preference relationship among the attributes. TCP-nets are defined as follows:
definition 3 (TCP-nets): TCP-nets is a graphical model G= < V, cp, ri, ci, cpt, cit >, where
(1)V＝{X 1 ,X 2 ,…,X n A set of nodes, corresponding to a set of qualitative service attributes;
(2) cp is a set of conditional preference arcs (cp-arc), cp-arc=<X i ,X j >Wherein X is i Is X j Conditional node of (1), i.e. X i ＝Pa(X j )；
(3) ri is a set of relative importance preference arcs (ri-arc), ri-arc=<X i ,X j >X represents i Ratio X j More importantly, i.e
(4) ci is a set of conditional relative importance arcs (ci-arc), ci-arc= (X) i ,X j ) X represents i And X j The importance of (2) depends on the value of the condition node;
(5) CPT is a set of Conditional Preference Tables (CPT), CPT (X) i ) For storing X under conditional nodes i A partial order relation of values;
(6) CIT is a set of Conditional Importance Tables (CIT), CIT (X) i ,X j ) For storing X under conditional nodes i And X j Is a partial order relationship of (a).
Fig. 2 is an example of TCP-nets. There are three different arc types in the figure, representing three preference relationships: the condition preference arc cp-arc marked with an arrow:<X 1 ,X 3 >which represents X 1 And X 3 A conditional preference relationship exists; importance preference arc ri-arc with black triangle labeling:<X 3 ,X 4 >which represents X 3 And X 4 There is a relative importance preference relationship; the ci-arc with black square marks is adopted (X) 2 ,X 3 ) Which represents X 2 And X 3 There is a conditional relative importance preference relationship. Accordingly, specific information of the preference relationship is illustrated by a Conditional Preference Table (CPT) and a Conditional Importance Table (CIT) in fig. 2.
Because a service has a plurality of qualitative service attributes, in order to screen out the best service, a preference model of qualitative service attribute combination of the service needs to be constructed on the basis of reasonable modeling representation of qualitative service attributes and preference thereof. The penalty value is adopted in this embodiment to quantitatively evaluate qualitative quality of service performance, and a specific method is described in detail in the following point 3. Thus, the service preference problem aims to select the service s that has the greatest utility in combination of qualitative and quantitative quality of service.
2. Consistency verification of qualitative preference relationships
Considering that conflict and redundancy situations may exist in preference information provided by a user, the embodiment designs a method for consistency check aiming at preference relation of qualitative service attributes.
According to definition 3, the preference relationship may be formally expressed as an arc in TCP-nets. Since the conditional preference has a higher priority than the relative importance preference and the conditional relative importance preference, if these two types of preference relationships exist simultaneously in two nodes and the directions are identical, the conditional preference relationship can be simplified. If the directions are inconsistent, the preference relationship of the two nodes is shown to have conflict, and the preference relationship of the two nodes cannot be generated in the TCP-nets. Based on the above analysis, the consistency check results can be classified into two types of compatibility and conflict, as shown in table 1. Gamma ray 1 And gamma 2 Represents qualitative service attributes X 1 And X 2 Two kinds of preference relations exist between the two kinds of 'gamma' 1 ⊙γ 2 "means the result of processing the preference relationship after consistency check. "operational description" is used to explain a specific processing scheme.
Table 1 qualitative service attributes node X 1 And X 2 Consistency determination table of preference relation between
The details of Table 1 are as follows:
(1) Compatibility: since condition preferences have higher priority than relative importance and condition relative importance, the latter can be included into the former relationship if they are in the same direction. Thus, if gamma 1 Is cp-arc, gamma 2 Is ri-arc or ci-arc, and their arc directions are the same, node X 1 And X 2 The preference relationship between them can be reduced to gamma 1 。
(2) Conflict: both the condition preference and the relative importance represent the dominant relationship between the two attributes. If gamma is 1 (i.e. X 1 Dominating X 2 ) And gamma 2 (i.e. X 2 Dominating X 1 ) At the same time, then node X 1 And X 2 There is a conflict in the preference relationship between them, requiring the user to re-describe.
In summary, redundant and conflicting information can be filtered by consistency checking the preference relationship of qualitative service attributes, thereby correctly describing the service preference relationship. After the consistency check is completed, TCP-nets can be generated so as to formalize the preference relation for further qualitative quantitative calculation.
3. Quantitative computation of qualitative service attribute combinations
The embodiment provides an improved penalty value calculation method for quantitatively evaluating qualitative service quality of service. Considering that the traditional penalty value calculation method can generate repeated calculation of the preference relationship, the embodiment adds an arc value matrix to quantize and store quantized values of the preference relationship, and uses node weights to preprocess each qualitative service attribute so as to facilitate quantized evaluation of qualitative service attribute combinations of subsequent specific services. The method can eliminate the problem of repeatedly calculating the preference relation in the traditional method.
(1) Calculating node weight and arc value
First, in order to represent the topological ordering of all nodes in a network graph G, it is necessary to generate a dependency graph G corresponding thereto. The dependency graph G corresponding to the TCP-nets network graph G of fig. 2 is shown in fig. 3. It can be seen that the dependency graph G is based on the network graph G, introducing a directed arc sci-arc:<X i ,X j >the pointing relationship of the conditional node and the corresponding node in ci-arc is shown. It should be noted that when generating the dependency graph G based on the network graph G, all ci-arc in the network graph G is found first, and for the corresponding nodes for constructing the ci-arc, each corresponding node shares a pre-node (i.e., a conditional node).
It can be seen that the dependency graph G gives nodes different degrees of importance because of the many preference relationships that exist in them. The present embodiment may characterize the importance of a node by its weight. Note that in G, the closer the current node is to the starting node, the greater its impact on the subsequent nodes (because the node at the front of the arrow represents a condition for which preference holds), and thus the greater the importance. Therefore, the embodiment can iteratively calculate the quantized values of each arc according to the reverse topological order, and determine the importance weight of the node according to the arc value quantization.
In addition, it has been found that the upstream arc value size depends on the arc value corresponding to the downstream node from which the preference relationship is generated, and that the same arc value may be used multiple times to calculate different upstream arc values. Therefore, to avoid repetitive calculations, the present embodiment introduces an arc value matrix Val to store quantized values for each preference relationship. The arc value matrix is defined as follows:
definition 4 (arc value matrix): the arc value matrix is val= [ v ] ij ] n×n Wherein n is the number of qualitative service attributes, v ij Is node X i And X j The value of the arc in between. Given two nodes X i And X j And an arc γ between nodes: (1) In the case of directional arcs (cp-arc, ri-arc, sci-arc), then v ij Storage Slave X i Pointing to X j And v is the arc value of ji =0; (2) If γ is an undirected arc (ci-arc), then v ij And v ji Respectively are provided withStore X i Ratio X j Important isAnd X j Ratio X i Important->Arc value in the case.
The iterative calculation of node weights and arc values is explained as follows:
for the node weight, it represents the importance of the node, and its weight is not less than the weight value of other nodes from the point. It can be determined by the maximum value of the arc from the node, the specific calculation being shown in equation (5).
Wherein ri (X) i )，ci(X i )，cp(X i ) And sci (X) i ) From X i Starting ri-arc, ci-arc, cp-arc and sci-arc sets.
The key to equation (5) is to calculate the arc value v ij Is of a size of (a) and (b). Influencing the arc value v ij There are two main factors of size. First, if attribute X j The more candidate values of (2) are, the more the attribute X is indicated i The more candidate combinations that are dominant, which means X i The greater the importance of (a) its arc value v ij And should be larger. Second, if attribute X j The greater the node weight due to X i Ratio X j More importantly, therefore its arc value v ij And will be larger. On the whole, the arc value v ij Is composed of X j Value range size and X of (2) j The node weights of (2) are determined together, and the specific calculation is shown in a formula (6).
v ij ＝w(X j )·|D(X j )| (6)
Wherein, |D (X) j ) I is X j Is a value range size, w (X) j ) Is X j Is a node weight of (c).
Due to the arc value v ij Only consider X i Ratio X j More important case, therefore for the undirected arc γ= (X) i ,X j ) For example, in calculating the arc value v ij Desired w (X) j ) Without taking into account the arc value v ji . Finally, it is emphasized that the arc values and node weights are iteratively calculated in reverse topological order. In reverse topology ordering, the first node has a node weight of 1 because it does not have any arcs. Then, according to the reverse topological order, the arc value and the node weight can be sequentially and iteratively calculated.
To sum up, the present embodiment describes how to generate an arc value matrix and a node weight list using algorithm 1. First, val storing the arc value and W storing the node weight are initialized, which are then used to calculate the penalty value (L1) for the particular service. Then, according to the reverse topological order of the dependency graph G, arc values related to each node are sequentially calculated, and weights (L2-L10) of the nodes are obtained. Specifically, different calculation methods (L3 to L8) are adopted for calculating the arc value according to the difference between the directional arc and the undirected arc. Finally, the node weight is calculated according to the quantized value of the arc starting from the node, so that the calculation of the related arc value and the weight of the upstream node is guided continuously (L9).
(2) Quantitative assessment of qualitative service attribute combinations
After the arc value matrix Val and the node weight list W are acquired, penalty value calculation can be performed on the combination of the values of the qualitative service attributes of the specific service s. Because of the qualitative preference relationship between attribute values and attributes, the calculation of penalty values is largely divided into two parts: rank (rank) s (X i ) Representing qualitative service attributes X in a particular service s i Value ordering of w s (X i ) Representing qualitative service attributes X in a particular service s i Importance weight of (c).
First, the present embodiment considers quantitative evaluation of each qualitative service attribute value in a specific service s, i.e., calculates rank s (X i ). Suppose qualitative service attribute X i The value range of (2) isThen rank is s (X i ) The value range of (2) is [0, |D (X) i )|-1]. If rank is s (X i ) The smaller the value, the better the attribute value. Taking the attributes and their preference relationships in fig. 2 as an example: for qualitative service attribute deployment platform (X) 2 ) The user prefers to a hybrid cloud, then a private cloud, and finally a public cloud. Then, if the deployment platform type of a particular service s is a hybrid cloud, then rank s (X 2 ) =0; if the cloud is public, rank s (X 2 )＝2。
Second, consider the weight of qualitative service attributes in a particular service, i.e., calculate w s (X i ). Note that in the calculation node weight and arc value section, the present embodiment has obtained the weight w (X i ). But w (X) i ) Only attribute node X is considered i More important than other nodes. For a particular service, if the attribute X i Absence of a conditional relative importance arc ci-arc, or ci-arc (X i ,X j ) All satisfy the requirement of the node X i Ratio X j Important condition, its weight attribute value w s (X i ) Can be directly equivalent to the weight value w (X) i ). Otherwise, in this embodiment, the weight w needs to be recalculated by using the arc value matrix in combination with the value of the specific service s and the relative importance relationship of the satisfied conditions s (X i ). To sum up, the weight attribute value w s (X i ) The calculation of (2) is shown in formula (7).
Wherein ri (X) i )，ci(X i )，cp(X i ) And sci (X) i ) From X i Starting ri-arc, ci-arc, cp-arc and sci-arc sets,representing attribute X i Specific attribute X j More important.
It can be seen that, in this embodiment, since the arc values of all preference relationships and the weight values of the nodes are stored in advance, only the node weights under the condition that the relative importance arcs of the conditions are not satisfied need to be calculated; and when calculating the attribute weight, the arc value matrix can be utilized to avoid a great deal of repeated calculation.
Finally, the present embodiment calculates the penalty value pen(s) for service s according to equation (8).
Where V is the qualitative service attribute set for service s.
In summary, a flowchart for calculating penalty values is shown in FIG. 4. First, the present embodiment generates a dependency graph G based on TCP-nets. Then, according to the inverse topological order of G, a node weight list W and an arc value matrix Val are obtained. Then, rank is calculated for a particular service s s (X i ) And simplify the pair W by W and Val s (X i ) To derive a penalty value pen(s).
Taking the dependency graph of fig. 3 as an example, the present embodiment first determines the arc value matrix Val and the node weight W according to algorithm 1. According to the reverse topological ordering of the nodes, the embodiment sequentially calculates X 4 、X 3 、X 2 、X 1 Is included, and node weights. Wherein due to node X 4 Arcs not pointing to other nodes, v 41 ＝v 42 ＝v 43 =0 and weight w (X 4 ) =1. Next, processing node X 3 . Due to the presence of the secondary X 3 Pointing to X 4 Relative importance arc of (c), thus v 34 ＝w(X 4 )·|X 4 |=1×2=2. In addition, X 3 、X 2 The existence of a conditional relative importance arc ci-arc between nodes, therefore X needs to be considered 3 Ratio X 2 Important situations. According to formula (6), in spite of X 2 Ratio X 3 In important cases, v 32 ＝w’(X 2 )·|X 2 |=1×3=3. To sum up, X 3 Is w (X) 3 ) =max (2, 3) +1=4. Then process node X 2 In spite of X 3 Ratio X 2 In important cases, v 23 ＝w’(X 3 )·|X 3 |=3×2=6, so X 2 Is w (X) 2 ) =6+1=7. Finally, according to v 12 ＝w(X 2 )·|X 2 |=7×3=21 and v 13 ＝w(X 3 )·|X 3 I=4×2=8, w (X 1 )＝21+8+1＝30。
Next, the present embodiment may evaluate the qualitative quality of each specific service by a penalty value. To serve s 1 For example, the qualitative service quality is "database Λ private cloud Λ china Λga". First, according to the preference relationship in FIG. 2, the present embodiment can calculate the rank values of the four attributes to be [0,1, 0 ]]. Due to X 1 Is a database, meaning X 3 Ratio X 2 More importantly, therefore, X 2 At s 1 The weight of (c) is updated to 0+1=1. And for X 1 、X 3 、X 4 For example, because the preference relationships remain true, their node weights are unchanged. To sum up, service s 1 The weight of each qualitative service attribute is [30,1,4,1 ]]. Finally, substituting the formula (8) to calculate the service s 1 Penalty value pen(s) 1 ) =5. Similarly, the present embodiment may calculate penalty values for other services, such as pen(s) 2 ) =1. Since the penalty value is smaller and better, the embodiment can obtain s according to the preference of the user 2 Is superior to s in qualitative service quality 1 。
It can be seen that, after the arc value matrix is introduced, the embodiment only needs to update the invalid ci-arc value according to the matrix, so as to obtain the specific weight of each attribute in the specific service, and does not need to perform iterative calculation again. Because the sorting of attribute values and the weight among the attributes are considered in the calculation of the penalty values, the method can keep consistent with the derivation result of the induced graph, is more suitable for the large service attribute combination problem, and can handle more complex preference problems. In addition, when the user only defines preference information between attribute values, that is, only CP-arc exists in the graph, the TCP-net is degenerated into CP-net, and the method of the embodiment is still applicable.
4. Service preference based on qualitative and quantitative service attributes
The penalty value of any service s can be obtained in this embodiment. To fuse with quantitative quality of service, this embodiment requires normalization of penalty values for the service. Considering that the larger the penalty value of a service, the worse the qualitative service quality of the service is indicated, the qualitative service attribute utility value Qual(s) of the service is shown in formula (9).
Where pen(s) represents the penalty value for service s. penn (penn) max Hepen min Is the largest penalty value and the smallest penalty value in all services.
Secondly, in terms of quantitative quality of service, the present embodiment uses classical aggregated utility value Quan(s) for evaluation, and since the calculation of the aggregated utility value Quan(s) of the quantitative service attribute belongs to the prior art, the present embodiment is not repeated. Finally, the integrated Utility value Utility(s) for the service s can be calculated by weighted summation of Qual(s) and Quan(s). And sorting the services according to the comprehensive utility value, and finding out the optimal service s.
5. Experimental analysis
Experimental setup
The present embodiment aims to evaluate qualitative service attributes based on an induction graph and on penalty values and preference processing methods thereof. Since the current QWS dataset records that web services contain only quantitative quality of service, there is no qualitative quality of service description. Thus, based on this dataset, the present embodiment randomly generates qualitative service attributes and their classification values, qualitative preferences, with reference to fig. 2. Considering that the processing of qualitative service attributes is the key for performing service preference, the present embodiment is compared with the processing method based on the induction map currently used by the improved penalty value calculation method proposed by the present embodiment for the qualitative processing section. All algorithms were implemented using python3.8, running on a machine with a 2.90GHz CPU and 16GB RAM.
Experimental results
(1) Experiment 1: changing the number of abstract tasks
In this set of experiments, the number of abstract tasks was 5-50, respectively. Each set of abstract tasks contains 500 candidate services. Each service includes randomly generated qualitative service attributes and qualitative preferences. The calculation time of the two qualitative service attribute solving methods is shown in fig. 5.
(2) Experiment 2: changing the number of candidate services
In this set of experiments, the number of candidate services per abstract task is 100-1000, respectively. The number of abstract tasks per group is fixed at 10. Each service includes randomly generated qualitative service attributes and qualitative preferences. The calculation time of the two qualitative service attribute solving methods is shown in fig. 6.
5-6, it can be seen that the penalty value based approach of the present invention uses less time than the inducement graph based approach for different numbers of tasks and different candidate services. And with the continuous increase of the number of tasks or the number of candidate services, the increase rate of the consumed time is far lower than that of the solution method based on the induction map.
Example two
The present embodiment provides a qualitative and quantitative hybrid cloud quality of service assessment system, comprising:
the acquisition module is used for: the cloud service preference relationship management method comprises the steps of acquiring qualitative service attributes of cloud services and preference relationships of the qualitative service attributes;
and (3) a checking module: the preference relation is used for carrying out consistency check on the preference relation of the qualitative service attribute to obtain a preference relation after consistency check;
modeling module: the method comprises the steps of modeling according to qualitative service attributes and preference relations after consistency inspection to obtain a graphical model;
and an evaluation module: the method comprises the steps of combining values of different qualitative service attributes based on the graphical model to obtain a specific service, and quantitatively evaluating the specific service;
the calculation module: the utility value of the qualitative service attribute is obtained through calculation according to the quantitative evaluation results of all the specific services;
and (3) optimizing module: and the utility value of the qualitative service attribute and the utility value of the known quantitative service attribute are weighted and summed to obtain a comprehensive utility value, and the optimal specific service is found out according to the comprehensive utility value.
While preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiments and all such alterations and modifications as fall within the scope of the application.
It is apparent that the above examples are given by way of illustration only and are not limiting of the embodiments. Other variations and modifications of the present invention will be apparent to those of ordinary skill in the art in light of the foregoing description. It is not necessary here nor is it exhaustive of all embodiments. And obvious variations or modifications thereof are contemplated as falling within the scope of the present invention.
Claims (10)
1. A qualitative and quantitative mixed cloud service quality assessment method is characterized in that: comprising the following steps:
step S1: acquiring qualitative service attributes and preference relations of cloud services;
step S2: consistency test is carried out on the preference relation of the qualitative service attribute to obtain a preference relation after consistency test;
step S3: modeling according to the qualitative service attribute and the preference relation after consistency test to obtain a graphical model;
step S4: based on the graphical model, combining the values of different qualitative service attributes to obtain a specific service, and quantitatively evaluating the specific service;
step S5: according to the quantitative evaluation results of all specific services, calculating to obtain utility values of qualitative service attributes;
step S6: and carrying out weighted summation on the utility value of the qualitative service attribute and the utility value of the known quantitative service attribute to obtain a comprehensive utility value, and finding out the optimal specific service according to the comprehensive utility value.
2. The qualitative and quantitative hybrid cloud service quality assessment method according to claim 1, wherein: the preference relation in the step S1 includes preference of qualitative service attribute value, and the preference of qualitative service attribute value is specifically divided into:
given a qualitative service attribute X i The value range isWherein (1)>Representing qualitative service attributes X i N of (2) i The number is further divided into two types according to whether preference conditions exist, including:
if no preference condition exists, unconditionally prefers UP k The formula is: wherein, > represents the partial order of attribute values;
if a preference condition exists, the CP is preferred for the condition k The formula is: CP (control program) k ＝[Pa(X i ):y]X i : Wherein [ Pa (X i ):y]The conditional attribute indicating the preference information is Pa (X i ) The conditional attribute takes the value y.
3. The qualitative and quantitative hybrid cloud service quality assessment method according to claim 1, wherein: the preference relation in the step S1 includes preferences between qualitative service attributes, and the preferences between qualitative service attributes are specifically divided into:
given two qualitative service attributes X i And X j Further, two categories are classified according to whether or not there is a preference condition, including:
if no preference condition exists, the relative importance RI k The formula is:wherein (1)>Representing a partial order between attributes;
if there is a preference condition, then it is the conditional relative importance CI k The formula is:wherein [ X ] c :y]The conditional attribute representing the relative importance is X c The conditional attribute takes the value y.
4. The qualitative and quantitative hybrid cloud service quality assessment method according to claim 1, wherein: in the step S3, modeling is performed according to the qualitative service attribute and the preference relationship after consistency test, so as to obtain a graphical model, and the method includes:
modeling qualitative service attributes and preference relations after consistency test through a graphical tool TCP-nets to obtain a graphical model G, wherein the graphical model G is expressed as:
G＝<V,cp,ri,ci,cpt,cit>
wherein v= { X 1 ,X 2 ,…,X n A set of nodes, corresponding to a set of qualitative service attributes; cp is a conditional preference arcSet of cp-arc, cp-arc=<X i ,X j >X represents i Is X j And X is the conditional node of (2) i ＝Pa(X j ) The method comprises the steps of carrying out a first treatment on the surface of the ri is a set of relative importance preference arcs ri-arc, ri-arc=<X i ,X j >X represents i Ratio X j More importantly, i.eci is a set of conditional relative importance arcs ci-arc, ci-arc= (X) i ,X j ) X represents i And X j The importance of (2) depends on the value of the condition node; CPT is a set of conditional preference tables CPT, CPT (X i ) For storing X under conditional nodes i A partial order relation of values; CIT is a set of conditional importance tables CIT, CIT (X i ,X j ) For storing X under conditional nodes i And X j Is a partial order relationship of (a).
5. The qualitative and quantitative hybrid cloud service quality assessment method according to claim 4, wherein: in the step S2, consistency checking is performed on the preference relationship of the qualitative service attribute, and the method includes:
let gamma 1 And gamma 2 Representing qualitative service attributes X 1 And X 2 Two preference relations, gamma 1 ⊙γ 2 Representing the processing result of the preference relation after consistency check, wherein the result of the consistency check is divided into two types of compatibility and conflict;
the compatibilizing comprises:
if gamma is 1 Is cp-arc:<X 1 ,X 2 >，γ 2 for ri-arc:<X 1 ,X 2 >gamma is then 1 ⊙γ 2 The result of (2) is gamma 1 ；
If gamma is 1 Is cp-arc:<X 2 ,X 1 >，γ 2 for ri-arc:<X 2 ,X 1 >gamma is then 1 ⊙γ 2 The result of (2) is gamma 1 ；
If gamma is 1 Is cp-arc:<X 1 ,X 2 >，γ 2 is ci-arc (X) 1 ,X 2 ) Gamma is then 1 ⊙γ 2 The result of (2) is gamma 1 ；
If gamma is 1 Is cp-arc:<X 2 ,X 1 >，γ 2 is ci-arc (X) 1 ,X 2 ) Gamma is then 1 ⊙γ 2 The result of (2) is gamma 1 ；
The conflict includes:
if gamma is 1 Is cp-arc:<X 1 ,X 2 >，γ 2 is cp-arc:<X 2 ,X 1 >gamma is then 1 ⊙γ 2 The result of (2) is None;
if gamma is 1 For ri-arc:<X 1 ,X 2 >，γ 2 for ri-arc:<X 2 ,X 1 >gamma is then 1 ⊙γ 2 The result of (2) is None;
if gamma is 1 Is cp-arc:<X 1 ,X 2 >，γ 2 for ri-arc:<X 2 ,X 1 >gamma is then 1 ⊙γ 2 The result of (2) is None;
if gamma is 1 For ri-arc:<X 1 ,X 2 >，γ 2 is cp-arc:<X 2 ,X 1 >gamma is then 1 ⊙γ 2 The result of (2) is None;
wherein None represents the need to be to node X 1 And X 2 And the preference relation between the two is described again until the preference relation is valid.
6. The qualitative and quantitative hybrid cloud service quality assessment method according to claim 5, wherein: based on the graphical model, combining the values of different qualitative service attributes to obtain a specific service, and quantitatively evaluating the specific service, wherein the method comprises the following steps:
introducing a directed arc sci-arc into the graphical model G:<X i ,X j >indicating the pointing relation between the conditional nodes and the corresponding nodes in the ci-arc to generate a dependency graph G which is used for indicating the topological ordering of all the nodes in the graphical model G;
according to the inverse topological order of the dependency graph G, a node weight list W and an arc value matrix Val are obtained, wherein the arc value matrix Val is used for storing quantized values of each preference relation, and the node weight list W comprises weight values W (X i )；
For a particular service s, calculate a qualitative service attribute X i Rank of values of (2) s (X i ) And simplifying qualitative service attribute X by using node weight list W and arc value matrix Val i Importance weight w of (2) s (X i ) To obtain a penalty value pen(s), wherein the specific service s is a service obtained by combining different values in different qualitative service attributes, and the qualitative service attributes X i Rank of values of (2) s (X i ) The definition is as follows: suppose qualitative service attribute X i The value range of (2) isRank is then s (X i ) The value range of (2) is [0, |D (X) i )|-1]。
7. The qualitative and quantitative hybrid cloud service quality assessment method according to claim 6, wherein:
the formula of the arc value matrix Val is as follows:
Val＝[v ij ] n×n
wherein n is the number of qualitative service attributes, v ij Is node X i And X j The value of the arc in between; given two nodes X i And X j And an arc gamma between nodes, v if gamma is any one of directional arcs cp-arc, ri-arc, sci-arc ij Storage Slave X i Pointing to X j And v is the arc value of ji =0; if gamma is the undirected arc ci-arc, then v ij And v ji Separately storeAnd->Arc value in case;
the weight value W (X i ) The formula is:
wherein ri (X) i )，ci(X i )，cp(X i ) And sci (X) i ) From X i Starting ri-arc, ci-arc, cp-arc and sci-arc sets;
v ij ＝w(X j )·D(X j )|
wherein, |D (X) j ) I is X j Is a value range size, w (X) j ) Is X j Is a node weight of (c).
8. The qualitative and quantitative hybrid cloud service quality assessment method according to claim 6, wherein: the penalty value pen(s) has the formula:
where V is the qualitative service attribute set of a particular service s, rank s (X i ) Representing qualitative service attributes X in a particular service s i Value ordering of w s (X i ) Representing qualitative service attributes X in a particular service s i The importance weights and formulas of (2) are:
wherein ri (X) i )，ci(X i )，cp(X i ) And sci (X) i ) From X i Starting ri-arc, ci-arc, cp-arc and sci-arc sets,representing attribute X i Specific attribute X j More important.
9. The qualitative and quantitative hybrid cloud service quality assessment method according to claim 1, wherein: in the step S5, according to the quantized evaluation results of all the specific services, a utility value of the qualitative service attribute is calculated, where the formula is:
where pen(s) represents the penalty value of a particular service s, pen max Hepen min Is the largest penalty value and the smallest penalty value in all specific services.
10. A qualitative and quantitative hybrid cloud quality of service assessment system, characterized by: comprising the following steps:
the acquisition module is used for: the cloud service preference relationship management method comprises the steps of acquiring qualitative service attributes of cloud services and preference relationships of the qualitative service attributes;
and (3) a checking module: the preference relation is used for carrying out consistency check on the preference relation of the qualitative service attribute to obtain a preference relation after consistency check;
modeling module: the method comprises the steps of modeling according to qualitative service attributes and preference relations after consistency inspection to obtain a graphical model;
and an evaluation module: the method comprises the steps of combining values of different qualitative service attributes based on the graphical model to obtain a specific service, and quantitatively evaluating the specific service;
the calculation module: the utility value of the qualitative service attribute is obtained through calculation according to the quantitative evaluation results of all the specific services;
and (3) optimizing module: and the utility value of the qualitative service attribute and the utility value of the known quantitative service attribute are weighted and summed to obtain a comprehensive utility value, and the optimal specific service is found out according to the comprehensive utility value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311263567.3A CN117336352A (en) | 2023-09-27 | 2023-09-27 | Qualitative and quantitative mixed cloud service quality assessment method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311263567.3A CN117336352A (en) | 2023-09-27 | 2023-09-27 | Qualitative and quantitative mixed cloud service quality assessment method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117336352A true CN117336352A (en) | 2024-01-02 |
Family
ID=89289644
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311263567.3A Pending CN117336352A (en) | 2023-09-27 | 2023-09-27 | Qualitative and quantitative mixed cloud service quality assessment method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117336352A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107609938A (en) * | 2017-09-07 | 2018-01-19 | 东南大学 | A kind of service recommendation method based on the qualitative and quantitative preference of user |
| CN109784722A (en) * | 2019-01-15 | 2019-05-21 | 齐鲁工业大学 | Web service selection method and system based on user preference |
| CN109871488A (en) * | 2019-02-22 | 2019-06-11 | 新疆大学 | A kind of Web service construction method and Web service for merging availability and user preference |
| US20200218579A1 (en) * | 2019-01-08 | 2020-07-09 | Hewlett Packard Enterprise Development Lp | Selecting a cloud service provider |
-
2023
- 2023-09-27 CN CN202311263567.3A patent/CN117336352A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107609938A (en) * | 2017-09-07 | 2018-01-19 | 东南大学 | A kind of service recommendation method based on the qualitative and quantitative preference of user |
| US20200218579A1 (en) * | 2019-01-08 | 2020-07-09 | Hewlett Packard Enterprise Development Lp | Selecting a cloud service provider |
| CN109784722A (en) * | 2019-01-15 | 2019-05-21 | 齐鲁工业大学 | Web service selection method and system based on user preference |
| CN109871488A (en) * | 2019-02-22 | 2019-06-11 | 新疆大学 | A kind of Web service construction method and Web service for merging availability and user preference |
Non-Patent Citations (1)
| Title |
|---|
| 徐九韵;武小丽;马竞峰;孙运雷;: "基于混合偏好的Web服务选择", 计算机技术与发展, no. 01, 10 January 2013 (2013-01-10), pages 139 - 146 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Garg | Hesitant Pythagorean fuzzy Maclaurin symmetric mean operators and its applications to multiattribute decision‐making process | |
| WO2021179834A1 (en) | Heterogeneous graph-based service processing method and device | |
| Zhou et al. | Group consistency and group decision making under uncertain probabilistic hesitant fuzzy preference environment | |
| Alfaro‐García et al. | Logarithmic aggregation operators and distance measures | |
| Wang et al. | Hesitant 2-tuple linguistic Bonferroni operators and their utilization in group decision making | |
| Durillo et al. | Convergence speed in multi‐objective metaheuristics: Efficiency criteria and empirical study | |
| Selvachandran et al. | A modified TOPSIS method based on vague parameterized vague soft sets and its application to supplier selection problems | |
| Yu et al. | Prioritized multi-criteria decision making based on preference relations | |
| Zou et al. | A new social network driven consensus reaching process for multi-criteria group decision making with probabilistic linguistic information | |
| Hirsch et al. | Variable preference modeling using multi-objective evolutionary algorithms | |
| Wang et al. | Predicting product co-consideration and market competitions for technology-driven product design: a network-based approach | |
| Agbodah | The determination of three-way decisions with decision-theoretic rough sets considering the loss function evaluated by multiple experts | |
| Lee et al. | Incremental analysis for generalized TODIM | |
| CN107122195A (en) | The software non-functional requirement evaluation method of subjective and objective fusion | |
| CN107203772B (en) | User type identification method and device | |
| Khezrian et al. | An approach for web service selection based on confidence level of decision maker | |
| Yang et al. | A multiple attribute group decision making approach for solving problems with the assessment of preference relations | |
| CN116628228B (en) | RPA flow recommendation method and computer readable storage medium | |
| Xian et al. | Interval probability hesitant fuzzy linguistic analytic hierarchy process and its application in talent selection | |
| WO2020085114A1 (en) | Information processing device, information processing method, and program | |
| Eslaminasab et al. | Determining appropriate weight for criteria in multi criteria group decision making problems using an Lp model and similarity measure | |
| CN117336352A (en) | Qualitative and quantitative mixed cloud service quality assessment method and system | |
| Beliakov et al. | Using linear programming for weights identification of generalized Bonferroni means in R | |
| Yang et al. | An attitudinal consensus method under uncertainty in 3PRLP selection | |
| Shankaranarayanan | Towards implementing total data quality management in a data warehouse |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |