CN101383854A - Web service providing method based on customer preference - Google Patents
Web service providing method based on customer preference Download PDFInfo
- Publication number
- CN101383854A CN101383854A CNA2008102256163A CN200810225616A CN101383854A CN 101383854 A CN101383854 A CN 101383854A CN A2008102256163 A CNA2008102256163 A CN A2008102256163A CN 200810225616 A CN200810225616 A CN 200810225616A CN 101383854 A CN101383854 A CN 101383854A
- Authority
- CN
- China
- Prior art keywords
- web service
- user
- service
- gang form
- web
- 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.)
- Granted
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a method for providing Web service based on a user preference. The method comprises the following steps: establishing a Web service composition template library which at least comprises a Web service composition template profile as well as a list and a process for abstract service, which are all described by an ontology concept in an ontology library; ensuring the preference degree of a user to the ontology concept; and searching a Web service composition template which is matched with a Web service request according to the Web service request of the user, if the Web service request is matched with the Web service composition template, the Web service composition template is provided to the user, otherwise, new service composition is constructed, and the constructed new service composition is provided to the user. The invention collects and analyzes preference information of the user in a quantized way and recommends the most suitable Web service to the user according to the obtained preference information of the user, therefore, the needs of the user are satisfied with the rapidest speed.
Description
Technical field
The method that provides of Web service is provided, the method that provides based on the user preference Web service is provided or rather.
Background technology
Along with the quick growth of quantity of service under the Web environment, how automatically according to service request, the dynamic combined of serving becomes a problem that presses for solution.Be directed to this, industrial quarters and academia have carried out big quantity research to the Web service combination from different perspectives, have proposed the various technology and the system of Web service combination.These technology can roughly be divided into following three classes according to the angle of methodology: based on artificial intelligence planning, based on workflow and based on software engineering method, below simply introduce it respectively.
For for the service combination of artificial intelligence planning, " Adapting Golog for Composition of Semantic Web Services " In Proceedingof the International Conference on the Principles of Knowledge Representation andReasoning referring to Sheila McIlraith and Tran CaoSon, 2002 and Sheila McIlraith, " SemanticWeb Services " In IEEE Intelligent Systems of Tran Cao Son and Honglei Zeng (Special Issue on the Semantic Web), 2001, a kind of formalization law of planning based on situation calculus is proposed, is knowledge representation formula about state and action, answer such problem with the deduction system of first order predicate calculus then: " if the state of certain objective attribute target attribute is satisfied in existence? exist, how can present state just be converted into dbjective state by action? "Answer to a such inquiry is plan that arrives dbjective state of structure, and this plan can be expressed as the sequence of action simply.Shortcoming based on the formalization law of planning of situation calculus is: the law of planning based on situation calculus is too abstract, is difficult to be applied in the middle of the real life.Simultaneously, the expression of knowledge also depends on the definition of state and action, and different definition can cause inconsistent result.
K.Erol, J.Hendler, and D.S.Nau etc. propose to utilize the descriptive language of OWL-S as Web service in " UMCP:A Sound and CompleteProcedure for hierarchical task-network planning (Proceedings of the InternationalConference on AI Planning Systems (AIPS); pp.249-254; June1994) ", and quote level task network (HTN in the artificial intelligence, Hierarchical Task Network) mode is set up the Web service combination by the method that task is decomposed.Different with other planing methods is, the core of HTN is not a state, but task.One is decomposed into the set of subtask based on the planning system of HTN with goal task, and these subtasks are resolved into littler subtask again, and this process that always circulates only comprises the atom Web service up to all subtasks, promptly up to decomposing the existing Web service.The shortcoming of this technical scheme is: simply the Web service combination being used as is decomposition to complex task, body gives expression to the Web service combination fully, and, do not consider user's preference information in the decomposable process, the Web service that is decomposited might not be satisfied user's demand.
On service combining method based on workflow, BPEL4WS (Business Process ExecutionLanguage for Web Services, BPEL) be a kind of important workflow compound mode, referring to BEA, IBM, Microsoft, SAP and Siebel., Business Process Execution Language forWeb Services, available at http://www-106.ibm.com/developerworks/web Services/library/ws-bpel 2003.BPEL combines Web service by a flow process, each step of flow process is called activity (activity), BPEL has defined atom action and structured activities control flow, defined partner (partner) and partner link (partnerLink) and be used for the Web service that each is different and include flow process in, the BPEL flow process is the centralized control point of Web service combination.Shortcoming: flexibility is not enough, needs predefined good berth flow model, and anabolic process do not have the support of semantic information, and the process of the web service is filled in abstract activity (Activity) in the execution phase belongs to the keyword coupling.
Under the mode based on soft project, model-driven is typical case's representative of this kind mode.The service combining method main thought of model-driven is that service groups is built jointly the formwork erection type, and model is constantly become specifically by abstract then, forms executable Web service combination process at last.Referring to Paolo Traverso, " the Automated Composition of Semantic Web Services into Executable Processes (the 3rd International Semantic Web Conference; 2004) " of Marco Pistore, the core of this technology is to use MBP (Model Based Planning) mode to carry out the combination of Web service, be characterized in the OWL-S description of existing Web service is changed into indecisive and the observable automaton of part, the Web service combination that generates at last also is a finite-state automata, and this finite-state automata can be converted to BPEL.Its shortcoming is: be not that all services all can be abstracted into the automaton form, thereby be difficult to constitute final finite state machine.
To sum up, in service combination field, industrial quarters lays particular emphasis on the language that proposes to describe the Web service combination at present, develops the related edit instrument and carries out engine, and academia then lays particular emphasis on from the Automatic Combined of aspect research Web services such as semantic reasoning, artificial intelligence planning.The technical method that industrial quarters proposes etc. has the good engineering exploitativeness, but exist the combination and exploitation difficulty greatly, aspects such as the system flexibility makeing mistakes easily, develop, extensibility have problems.Therefore, providing the Web service combined aspects to the user, is not very complete, especially, the Web service combination that is provided needs and can meet consumers' demand, and the demand between the different user there are differences, and present technology all can not provide its required Web service combination to the user well.
Summary of the invention
In view of this, main purpose of the present invention is to provide the method that provides based on the user preference Web service, the Web service combination is quantized, and determine required Web service combination as the user, maximum possible satisfies user's demand, provides the efficient of Web service combination also will be promoted greatly.
For achieving the above object, technical scheme of the present invention is achieved in that
A kind of method that provides based on the user preference Web service is set up Web service gang form storehouse, and described Web service gang form comprises at least: Web service gang form profile is used to describe the essential information of Web service gang form; The tabulation of abstract service is used to describe abstract Web service; Process is used to describe logical relation and execution sequence between the abstract Web service; Wherein, the tabulation of described Web service gang form profile and abstract service is all described by the Ontological concept in the ontology library; Recording user for each Ontological concept is set up user preference table, is determined user preference to this Ontological concept according to described degree of recognition to the degree of recognition of Ontological concept in user and system interaction process; This method comprises:
After receiving the Web service request, utilize the formalized description of described Web service gang form, in described Web service gang form storehouse, search Web service gang form with described Web service request coupling, if have the Web service gang form that mates fully with the Web service request, then described Web service gang form be back to the user; If do not have the Web service gang form that mates with the Web service request, then make up new Web service gang form:
Search each Web service with described Web service request coupling, in each Web service that is found out, choose the Web service of user preference degree maximum respectively by user's request, selected Web service is recommended to the user, all the other Web services are saved in the stack, if the Web service that user's acceptance is recommended, then as the Web service that makes up the Web service gang form, if the user does not accept the Web service recommended, then in described stack, carry out retrospective search, again with retrospective search to Web service recommend to the user; Determine whether each Web service need introduce new solicited message in the new Web service gang form of creating, and when having new solicited message, searches the Web service of satisfying this new solicited message, and recommends to the user; Retrospective search failure and new solicited message are searched when finishing, and will be made up the Web service gang form by the Web service that the user accepts, if the Web service of not accepted by the user, then the Web service gang form makes up failure;
If there is the Web service gang form that not exclusively mates with the Web service request, then based on the Web service gang form of described incomplete coupling, making up provides to the Web service gang form:
Do not search the Web service of coupling for there being request item in the Web service request that Web service matches, in each Web service that is found out, choose the Web service of user preference degree maximum respectively by user's request, selected Web service is recommended to the user, all the other Web services are saved in the stack, if the Web service that user's acceptance is recommended, then as the additional Web service of the Web service gang form of described incomplete coupling, if the user does not accept the Web service recommended, then in described stack, carry out retrospective search, again with retrospective search to Web service recommend to the user; Determine that whether described additional Web service need introduce new solicited message, when having new solicited message, searches the Web service of satisfying this new solicited message, and recommends to the user; Retrospective search failure and new solicited message are searched when finishing, and the Web service that will be accepted by the user makes up new Web service gang form as the additional Web clothes of the Web service gang form of described incomplete coupling;
Described Web service gang form is provided to the user, and for the Web service gang form of new structure, if accepted by the user, then the Web service gang form with described new structure is added into described Web service gang form storehouse.
Wherein, described Web service specifically describes and is: and WebService (WSInput, WSOutput, WSPrecondition, WSEffect), wherein, WebService is the title of Web service; WSInput is the input information set of service; WSOutput is the output information set of service; WSPrecondition is the precondition of service; WSEffect is the influence that service produces; Service request is described concrete structure: and RequestWS (RWSInput, RWSOutput), wherein, RequestWS is the title of Web service request; RWSInput is the input information set that the user provides; RWSOutput is the target information set of user's service request; Everyly in described Web service, the Web service request all describe by the Ontological concept in the ontology library.
Wherein, described Web service gang form profile specifically describes: TemplateProfile (TemplateInput, TemplateOutput, TemplateCategory, Precondition, Effect), wherein, TemplateInput is the input information of described Web service gang form, corresponding more than one Ontological concept; TemplateOutput is the output information of described Web service gang form, also corresponding more than one Ontological concept; TemplateCategory represents the classified information of described Web service gang form; Precondition represents the precondition of described Web service gang form; Effect represents effect and the influence after described Web service gang form is carried out; The tabulation of described Web abstract service comprises more than one Web abstract service, described Web abstract service specifically describes: AbstractWS (AbstractWSInput, AbstractWSOutput, Precondition, Effect), wherein, AbstractWSInput is the input information of described Web abstract service, corresponding more than one Ontological concept; AbstractWSOutput is the output information of described Web abstract service, also corresponding more than one Ontological concept; AbstractWSCategory is the classified information of described Web abstract service; Precondition is the precondition of described Web abstract service; Effect is effect and the influence after described Web abstract service is carried out.
Wherein, the user preference degree of described Ontological concept is pressed following formula and is determined:
Wherein, N=(Ontological concept is as the number of times of user's service request input information) * a%+ (Ontological concept is as the number of times of user's service request output information) * b%+ (number of times that number of times * 2-of being accepted by the user in recommendation process of Ontological concept is refused by the user) * c%; Wherein, the value of a is 10 to 20, and the value of b is 30 to 40, and the value of c is 50 to 60;
Matching degree between described Web service gang form and the user's Web service request RequestWS is pressed following formula and is determined:
Wherein, the matching degree of the notion of mating most in the set of the input information of the matching degree of RWSInputi notion=RWSInputi notion and Web composite services template; The matching degree of the notion of mating most in the output information set of the matching degree of RWSOutputj notion=RWSOutputj notion and Web composite services template.
Wherein, the function of determining the user preference of Web service is:
f(n)＝g(n)+h(n)
Wherein,
The Input notion number of n for not provided by the user at present in the Web service, Inputi is provided by n the Ontological concept in i the parameter in the Input information that can not be provided by the user.
Wherein, described logical relation comprises order, parallel and selection at least.
Web service of the present invention is described by the semantic Ontological concept in the field ontology library, after receiving the Web service request, by quantizing the preference information of collection and analysis user, and according to resulting user's preference information, recommend only Web service to the user, the demand that satisfies the user of prestissimo.In addition, also greatly improved the reusability of Web service gang form, made Web service be combined on the basis of existing template support and need not from reconfiguring separately based on the administrative skill of template, improved system the Web service combination efficient is provided.
Description of drawings
Fig. 1 is the flow chart that method is provided based on the user preference Web service of the embodiment of the invention;
Fig. 2 is the structural representation of the composite services template T1 of the embodiment of the invention;
The Web service combinatorial topology figure that Fig. 3 obtains after for the request of Web service first of the embodiment of the invention;
The Web service combinatorial topology figure that Fig. 4 obtains after for the request of Web service once more of the embodiment of the invention.
Embodiment
Below in conjunction with accompanying drawing technical scheme of the present invention is elaborated.
Among the present invention, the Web service describing mode is specially: and WebService (WSInput, WSOutput, WSPrecondition, WSEffect), wherein, WebService is the title of service, and each Web service all has title, and the title of Web service can be identical; WSInput is the input information set of Web service, and these input informations are all described by the Ontological concept in the semantic ontology library; WSOutput is the set of the output information of Web service, and is same, and output information is also described by the Ontological concept in the semantic ontology library; WSPrecondition is the precondition of Web service, is used to limit Web service; WSEffect is the influence that Web service produces, i.e. the issuable effect of Web service.In the embodiments of the invention, lay particular emphasis on application to WSInputSet parameter and the WSOutputSet parameter of WebService, the description of Web service can be reduced to WebService (WSInput, WSOutput).User's service request describing mode of the present invention is: and RequestWS (RWSInput, RWSOutput), wherein RequestWS is the name of Web service request; RWSInput is the input information set that the user can provide; RWSOutput is the target information set of user's service request.Here, the parameter information of RWSInput and RWSOutput is described by the Ontological concept in the semantic ontology library equally.
Ontology originates from the Western Philosophy in ancient times, from the eighties of last century the eighties, ontological thought is introduced into artificial intelligence field, and people begin progressively real world to be carried out modeling, from certain field, take out the set of notion and the relationship of the concepts, promptly form the body in this field.Body helps to change different groups or system adopts the situation of different term descriptions to identical concept in the field, thereby is widely used in numerous areas such as digital library, information integration.The advantage that adopts Ontological concept in the semantic ontology library to describe Web service is, described object has higher unified standard, described object is standard more, and, because semantic ontology library self characteristics, correlation between the described Web service is easier determines that the semantic features of Web service can obtain performance biglyyer.Because the correlation technique of semantic body is a known technology, neither emphasis of the present invention, therefore repeat no more here.
Fig. 1 is the flow chart that method is provided based on the user preference Web service of the embodiment of the invention, and as shown in Figure 1, the method that provides based on the user preference Web service of the embodiment of the invention may further comprise the steps:
Step 101: set up Web service gang form (WSTemplate) storehouse, described Web service gang form comprises at least: Web service gang form profile (TemplateProfile) is used to describe the essential information of Web service gang form; The tabulation of abstract service (AbstractWSList) is used to describe abstract Web service; Process (Process) is used to describe logical relation and execution sequence between the abstract Web service, and the tabulation of described Web service gang form profile and abstract service is all described by the Ontological concept in the ontology library; Recording user for each Ontological concept is set up user preference table, is determined user preference to this Ontological concept according to described degree of recognition to the degree of recognition of Ontological concept in user and system interaction process.
Store a large amount of Web service combinations in the Web service gang form storehouse, these Web service combinations form by artificial the setting automatically, or generation automatically by user's use habit, in a single day certain or some Web service combinations are defined as the Web service combination that the user uses always, just it is stored in the Web service gang form storehouse as template, the Web service gang form is normalized Web service combination.
Web service gang form WSTemplate is described as: (Process), wherein, WSTemplate represents the title of Web service gang form to WSTemplate=for TemplateProfile, AbstractWSList, and each Web service combination all has corresponding title.TemplateProfile=(TemplateInput, TemplateOutput, TemplateCategory, Precondition, Effect), the essential information of TemplateProfile description template, wherein, TemplateInput is the input information of template, is described by one or more Ontological concept; TemplateOutput is the output information of template, is also described by one or more Ontological concept; TemplateCategory represents the classified information of these composite services; Precondition is the precondition of template; Effect is effect and the influence after template is carried out.AbstractWSList describes the tabulation of abstract service, and one or more abstract service can be arranged.Each abstract service can be by a quadruple notation, and AbstractWS=(AbstractWSInput, AbstractWSOutput, Precondition, Effect).Wherein AbstractWSInput is the input information of abstract service, is described by one or more Ontological concept; AbstractWSOutput is the output information of abstract service, is also described by one or more Ontological concept; AbstractWSCategory represents the classified information of abstract service; Precondition is the precondition of abstract service; Effect is effect and the influence after abstract service is carried out.The structure collection of Process description template, logical relation and execution sequence between each abstract service that comprises in the expression template, logical relation comprises order (Sequence), parallel (Parellel) and selects (Choice) three kinds.Below describe respectively:
(Sequence) is the binary crelation on the AbstractWSList in proper order:
Sequence={＜AbstractWS
i, AbstractWS
j| AbstractWS
i, AbstractWS
j∈ AbstractWSList, and if only if AbstractWS
iAfter the end, AbstractWS
jCould begin to carry out }
Parallel (Parellel) is the binary crelation on the AbstractWSList:
Parellel={＜AbstractWS
i, AbstractWS
j| AbstractWS
i, AbstractWS
j∈ AbstractWSList, and if only if AbstractWS
iAnd AbstractWS
jCan executed in parallel }
Selection (Choice) is the binary crelation on the AbstractWSList:
Choice={＜AbstractWS
i, AbstractWS
j| AbstractWS
i, AbstractWS
j∈ AbstractWSList, and if only if AbstractWS
iAnd AbstractWS
jBetween have only an abstract service to be performed.
Utilize the formalized description definition of Web service gang form, can obtain the matching degree between Web service gang form and the user request easily.Below will elaborate.
The user preference degree of Ontological concept is pressed following formula and is determined:
Wherein, N=(Ontological concept is as the number of times of user's service request input information) * a%+ (Ontological concept is as the number of times of user's service request output information) * b%+ (number of times that number of times * 2-of being accepted by the user in recommendation process of Ontological concept is refused by the user) * c%; Wherein, the value of a is 10 to 20, and the value of b is 30 to 40, and the value of c is 50 to 60.
Owing to the user preference degree of Ontological concept recommends the increase of number of times to increase along with the user accepts this Ontological concept, and the value of user preference degree is along with the continuous analysis to user profile, finally should form comparison smooth variation curve, therefore select the basic function of logarithmic function log (N+1) as Ontological concept user preference degree.When the value of N+1 between (1,10) time, the span of Ontological concept user preference degree is between (0,1); When the value of N greater than 10 the time, Ontological concept user preference degree is greater than 1.Choosing of above-mentioned weights, mainly be divided into two parts, a part is an Ontological concept in the input of user's request, the output message embodiment to preference, and is starting point with the customer objective, obviously, the information of output has stronger purpose than input information, and input information may be undistinct in the starting stage, for example may be user's fuzzy query, but information as output, be the purpose that user expectation reaches, always very clear and definite, therefore give bigger weights to the information of output.Another part is that the weights in the Ontological concept recommendation process are given, the difference of the number of times of the number of times that the user accepts and user's refusal can reflect the preference degree of user to certain Ontological concept, and the preference meaning of accepting Ontological concept fully and being reflected is greater than the refusal meaning that a time notion reflected, the present invention adopts the user to accept 2 times of number of times to carry out difference with the refusal number of times, the shared weight of Ontological concept recommendation process and occupy bigger proportion than having Ontological concept input and output message in the request.The setting of above-mentioned weighted value has taken into full account user's demand, and therefore the result who is obtained has more hommization.User's demand is quantized in the embodiment of the invention as far as possible, guaranteed the result's that obtains unification, guarantee to satisfy requirements of different users as much as possible under the prerequisite of science.
Matching relationship between the Ontological concept is specially:
The matching degree of Ontological concept A and Ontological concept B is:
Because the matching degree between the Ontological concept is not an emphasis of the present invention, so the present invention has only considered that in order to simplify the matching algorithm of Ontological concept A and B are the situations of set membership and brotherhood.Set membership and brotherhood are typical matching relationships, have also simplified the realization difficulty of technical solution of the present invention simultaneously.
When the embodiment of the invention makes up in the generation Web service, recording user and user's intersection record, and with this foundation as the user preference degree.
Step 102: after receiving the Web service request, utilize the formalized description of described Web service gang form, in described Web service gang form storehouse, search Web service gang form with described Web service request coupling, if have the Web service gang form that mates fully with the Web service request, then described Web service gang form be back to the user; If do not have the Web service gang form that mates with the Web service request, then make up new Web service gang form:
Search each Web service with described Web service request coupling, in each Web service that is found out, choose the Web service of user preference degree maximum respectively by user's request, selected Web service is recommended to the user, all the other Web services are saved in the stack, if the Web service that user's acceptance is recommended, then as the Web service that makes up the Web service gang form, if the user does not accept the Web service recommended, then in described stack, carry out retrospective search, again with retrospective search to Web service recommend to the user; Determine whether each Web service need introduce new solicited message in the new Web service gang form of creating, and when having new solicited message, searches the Web service of satisfying this new solicited message, and recommends to the user; Retrospective search failure and new solicited message are searched when finishing, and will be made up the Web service gang form by the Web service that the user accepts, if the Web service of not accepted by the user, then the Web service gang form makes up failure;
If there is the Web service gang form that not exclusively mates with the Web service request, then based on the Web service gang form of described incomplete coupling, making up provides to the Web service gang form:
Do not search the Web service of coupling for there being request item in the Web service request that Web service matches, in each Web service that is found out, choose the Web service of user preference degree maximum respectively by user's request, selected Web service is recommended to the user, all the other Web services are saved in the stack, if the Web service that user's acceptance is recommended, then as the additional Web service of the Web service gang form of described incomplete coupling, if the user does not accept the Web service recommended, then in described stack, carry out retrospective search, again with retrospective search to Web service recommend to the user; Determine that whether described additional Web service need introduce new solicited message, when having new solicited message, searches the Web service of satisfying this new solicited message, and recommends to the user; Retrospective search failure and new solicited message are searched when finishing, and the Web service that will be accepted by the user makes up new Web service gang form as the additional Web clothes of the Web service gang form of described incomplete coupling.
Below describing the present invention in detail is how to realize that the Web service combination makes up.
Utilize the formalized description definition of Web service gang form, can obtain the matching degree between Web service gang form and the user's Web service request.
Web service gang form WSTemplate and user ask the matching degree of RequestWS to be described as among the present invention:
Wherein, the matching degree of the Ontological concept that mates most in the set of the input information of the matching degree of RWSInputi notion=RWSInputi notion and Web composite services template; The matching degree of the Ontological concept that mates most in the output information set of the matching degree of RWSOutputj notion=RWSOutputj notion and Web composite services template.
The function of determining the user preference of Web service is:
f(n)＝g(n)+h(n)
Wherein,
The Input notion number of n for not provided by the user at present in the Web service, Inputi is provided by n the Ontological concept in i the parameter in the Input information that can not be provided by the user.
The present invention considers when candidate's Web service is selected, need on the one hand relatively Web service as the candidate add service make up after and the deviation between user's service request, g in the formula (n) is exactly the calculating to this value.The user preference degree of the Input notion that can not be provided by the user is provided to analyze in candidate's Web service at present on the other hand, thereby analyzes the possibility that the user provides these Ontological concepts, h in the formula (n) is exactly the calculating to this value.Therefore, this function can be made analysis according to user's historical information in the Web service anabolic process, and the Web service that will meet user's request most recommends among the composite services.
Describe g (n) and h (n) function below in detail:
The implication of g (n) is: at present can not be many more in the Web service by the Input notion number that the user provides, show that the distance of this Web service and user's service request is just far away more, and therefore, g (n) must be a monotonic decreasing function.And, along with the increase of n, candidate's Web service add after the service combination and the deviation between user's service request just big more, g (n) just levels off to 0 more, considers actual application, with
For the exponential function at the end more can embody above-mentioned relation, greater than 0 o'clock, value was the decreasing function on (0,1) to this function at n.
The implication of h (n) is: a plurality of Input notions that can not be provided by the user may be provided in the Web service, and therefore, the mean value of the user preference degree of these Input notions that can not be provided by the user is provided h (n).Have only when these Input notions exist higher user preference degree higher, h (n) just can be greater than 1.
In sum, among formula f (n)=g (n)+h (n), the shared proportion of g (n) and h (n) is basic identical, but when the user is higher to the preference of some Ontological concept, the proportion of h (n) will increase, and this also proposes to provide a basis based on the heuristic Web service combined strategy of user preference for the present invention.
About step 102, specifically may further comprise the steps:
Step 1: after receiving the Web service request of user's input, search corresponding template at ATL.
About the Web service request, can come standard user's input by certain qualification form, as provide the corresponding Ontological concept in the semantic ontology library to supply user's selection etc., corresponding pattern of the input can be determined according to the needs of practical application.
Step 2: according to aforesaid matching degree computing formula, calculate the matching degree of Web service gang form, select the highest template of matching degree as optimal Template.If the optimal Template matching degree equals 1, there is the service moulding plate of coupling fully in expression, and then Web service is made up successfully, directly returns the Web service gang form that is found out, and changes step 13.If Web service gang form matching degree equals 0, all Web service gang forms are all inconsistent with user's request in the expression ATL, need rebuild the Web service combinational logic, and it is 0 that the modified logo position is set, to rebuild template but not revise, enter step 3.If less than 1, the expression template part meets user's request to the template matches degree greater than 0, the template of part coupling is preserved, being provided with and revising the expression position is 1, is illustrated in the later calculating iterative process, revises combinational logic on this template basis, enters step 3.
Step 3: Web service is searched in request according to Web service, and output information is decided to be target, finds out the Web service that can satisfy these output informations.In the process that iteration is searched, the target of inquiry is the output information in user's service request for the first time.Carry out in the process at algorithm, search target and be the new input information of introducing of service.
Step 4: if there is not the set of service that finds out, illustrate in this iterative process that target output information does not have Web service to provide in network, change step 6.If there is set of service, enter step 5.Here, set of service is to satisfy all Web service set that found out of a certain Web service request.
Step 5: the user preference degree according to Ontological concept is selected the Web service in the set of service, calculate the user preference of each Web service, and according to a Web service selecting the preference maximum based on the heuristic function of user preference, as the optimal service of next step Web service combination.And all the other not selected Web services are saved in the stack, change step 7.
Step 6: Web composite services process is recalled, whether had Web service in the inquiry stack.If not having Web service in the stack is that stack is empty, then change step 14; If stack is not empty, get stack top element as alternative Web service, promptly get the Web service of mating most, change step 8; It will be appreciated by those skilled in the art that recalling is known technology in this area, its many common technology means of searching as parameter repeat no more here.
Step 7: investigate optimal service and whether introduce new input information, new input information refers to: the information that can not provide in information that import in user's request, output does not all provide and the template to be revised.If there is new input information, change step 9:, enter rapid 8 if there is not new input information.
Step 8: optimal service is not introduced new input information, then finishes the service combination.Change step 13.
Step 9: optimal service is introduced new input information, then carries out alternately with the user, determines whether new input information user accepts.Enter rapid 10.
Step 10: obtain user feedback message, divide three kinds of situations:
1, the user does not accept the service recommended fully, and directly refusal enters rapid 11.
2, the user accepts recommendation service, but required input can not be provided, and changes step 3.
3, the user accepts recommendation service, and required input information can be provided.Change step 12.
Step 11: the user refuses this recommendation service, reduces relevant Ontological concept user preference degree.Change step 6.
Step 12: the user accepts recommendation service, improves relevant Ontological concept user preference degree, enters rapid 13.
Step 13: service groups is synthesized merit, and according to the modified logo position, structure or the final service of modification gang form deposit ATL in, so that be used in the anabolic process later on.
Step 14: service combination failure, can't provide the user required service in the current system, return corresponding failed message to the user.
The effect that the present invention is based on the user preference heuristic function is the interactive information according to user's service request and user and system, for the user recommends suitable service.Therefore, admissible parameter has the input information and the output information of user's service request; When user and system interaction, to the approval number of times of Ontological concept.By these information, find that the user has special preference to some notion, therefore, the present invention has set up the Ontological concept preference table for each Ontological concept, the user preference degree of each Ontological concept corresponding these notions of value representation in table, initialization when each user's Ontological concept preference value is registered by the user, with the reciprocal process of system in, system will be according to the reciprocal process between user and the system, automatically analysis and definite user are to the preference degree of Ontological concept, concrete preference account form is referring to aforesaid each formula.Different users embodies different user preferences.The reciprocal process that it will be appreciated by those skilled in the art that recording user and system is to realize easily, like this, according to the computing formula of parameter that is write down and the aforesaid preference of substitution, can determine the preference of each user to Ontological concept easily.Need to prove that Web service is also described by input information and output information,, just can calculate the user preference degree of Web service so according to the user preference degree of the relevant Ontological concept of these information.
It will be appreciated by those skilled in the art that after having determined to make up each Web service of Web service combination,, directly generate relevant Web service combination and get final product by the structure form of Web service combination.
It is emphasized that in the embodiment of the invention that determined formula is according to user's actual conditions and definite analog function, these functions all have actual directive significance.
Step 103: Web service gang form constructed in the step 102 is provided to the user, and for the Web service gang form of new structure, if accepted by the user, then the Web service gang form with described new structure is added into described Web service gang form storehouse.
In the step 103, export the Web service combination to for each, after the user accepts fully, being about to it is added in the described Web service gang form storehouse as the Web service gang form, so that after the user will import identical Web service request from now on, directly in Web service gang form storehouse, find the Web service combination of this new structure that meets user's request.
Below, illustrate essence of the present invention once more by an example.
The present invention is example with the tour field, create the set of all bodies of tourism Ontological concept set G, Web service combination, body such as tour field: air ticket (Airline Ticket), hotel (Hotel), the city (Origination City) of setting out, target cities (Destination City), these all are by the notion in the ontology library of domain expert's definition, ontology library uses the OWL definition.Create certain user's Ontological concept preference correspondence table, see the following form 1:
| Notion | The user preference degree |
| Origination?City | 0 |
| Travel?Sight | 0 |
| Date | 0 |
| Destination?City | 0 |
| Hotel | 0 |
| Airline?Ticket | 0.2 |
| Train?Ticket | 0.1 |
| Ticket | 0 |
| Ticket?Info | 0 |
| Travel | 0 |
Table 1
And this user to Airline Ticket Ontological concept, the user preference degree of Train Ticket Ontological concept is initialized as 0.2 and 0.1, the user preference degree of other notions is initialized as 0.Utilize the Ontological concept of these tour fields, define relevant Web service by the domain expert, as table 2:
| The Web service title | The Web service input information | Web service output information | The Web service classified information |
| WS1 | Travel?Sight | Destination?City | Travel?Service |
| WS2 | Origination?City， Destination?City， Date | Train?Ticket | Travel?Service |
| WS3 | Destination?City， Date | Hotel | Travel?Service |
| WS4 | Origination?City， Destination?City， Date | Airline?Ticket | Travel?Service |
| WS5 | Hotel， Train?Ticket | Travel | Travel?Service |
| WS6 | Hotel， Airline?Ticket | Travel | Travel?Service |
| WS7 | Ticket | Ticket?Info | Travel?Service |
Table 2
And Web composite services template T1, as shown in Figure 2, tourist attractions (Travel Site), sailing date tourism Ontological concept information such as (Date) represent with circle; Abstract Web service is represented with rectangle in the template; Input is represented with arrow; The template output set is represented with the solid line oval frame; The template input set is share the dotted ellipse frame table and is shown.
The present invention will pass through twice Web service request Benq of user respectively in the heuristic Web service combined strategy of user preference.
User's service request for the first time:
User's target is to obtain nonlocal travel information (Travel), the user can comprise the city of setting out (Origination City), tourist attractions (Travel Site) and sailing date (Date) to the information that system provides, description definition according to aforementioned ServiceRequest, user's service request can be expressed as ServiceRequest (Input (Origination City, Travel Site, Date), Output (Travel)).
According to user's service request information ServiceRequest and existing Web service in the system at present, obtain the Web service topological diagram of this request, as shown in Figure 3.
After receiving the Web service request, handle by following process:
Step 1a: accept user's request ServiceRequest, and in ATL, inquire about only template.
Step 2a: the template T1 that mates most in the ATL can not satisfy user's service request fully, according to aforesaid matching degree computing formula, the matching degree that can calculate template T1 and user's service request is 0.5, will make amendment on the basis of the Web service gang form that is found.
Step 3a: as querying condition, searching output information can be provided is Web service WS5 and the WS6 of Travel with user's output information Travel.
Step 4a: calculate the degree that WS5 and these two services of WS6 are subjected to user preference according to the user preference heuristic function, and the service recommendation that preference heuristic function value is the highest is given the user.The new input information that does not provide among user's input information and the template T1---Train Ticket and AirlineTicket all are provided for WS5 and WS6.
Step 5a:WS6 is because introduced new input information Airline Ticket, the information that all can not provide for user and template, thereby need carry out alternately with the user, system offers the user with WS6, and whether the inquiry user can provide the relevant information of Airline Ticket among the WS6.
Step 6a: WS6 is accepted in user feedback, but Airline Ticket can not be provided relevant information, and system will continue to search Web service with Airline Ticket as querying condition.And according to Ontological concept user preference degree
Computing formula, the user preference degree of Ontological concept Airline Ticket information is revised.The user preference of Airline Ticket: log (N+1)=log (((2-0) * 50%)+1)=0.30.New user preference value is saved in the user preference table.
Step 7a: search all Web services as output with Airline Ticket, find out WS4, the Airline Ticket information that it both can obtain exporting, satisfy the input information of the OriginalCity that user request provided simultaneously, thereby the request to the user is satisfied fully, therefore in conjunction with the template that had before found, constructs a Web service combination chain, as shown in Figure 4, the result with the Web service combination presents to the user.And adjust the user preference information of all the other Ontological concepts.As the input Ontological concept Origination City that in user's request, occurs, its user preference information log (N+1)=log (1 * 15%+1)=0.06.Output Ontological concept Travel, its user preference information=log (1 * 35%+1)=0.13.
Step 8a: the template that will newly make up deposits in the ATL, so that similar service is provided in the later query script.
After once asking, the user preference degree of the related notion in this user request information all gets a promotion, and the notion preference of output promotes the preference greater than the input notion.In reciprocal process, the user accepts WS6 and serves needed input information Airline Ticket, although the user can't provide Airline Ticket the relevant information of notion, but still accepted this notion, show that the user is still to this notion preference to some extent, therefore the calculating of the preference of the present invention's proposition makes its notion user preference degree strengthen to some extent, and table 3 is adjusted user preference degree correspondence table:
| Notion | The user preference degree |
| Origination?City | 0.06 |
| Travel?Sight | 0.06 |
| Date | 0.06 |
| Destination?City | 0 |
| Hotel | 0 |
| Airline?Ticket | 0.3 |
| Train?Ticket | 0.1 |
| Ticket | 0 |
| Ticket?Info | 0 |
| Travel | 0.13 |
Table 3
Web service of the present invention is described by the semantic Ontological concept in the field ontology library, after receiving the Web service request, by quantizing the preference information of collection and analysis user, and according to resulting user's preference information, recommend only Web service to the user, the demand that satisfies the user of prestissimo.In addition, also greatly improved the reusability of Web service gang form, made Web service be combined on the basis of existing template support and need not from reconfiguring separately based on the administrative skill of template, improved system the Web service combination efficient is provided.
The above is preferred embodiment of the present invention only, is not to be used to limit protection scope of the present invention.
Claims (6)
1, a kind of method that provides based on the user preference Web service is characterized in that, sets up Web service gang form storehouse, and described Web service gang form comprises at least: Web service gang form profile is used to describe the essential information of Web service gang form; The tabulation of abstract service is used to describe abstract Web service; Process is used to describe logical relation and execution sequence between the abstract Web service; Wherein, the tabulation of described Web service gang form profile and abstract service is all described by the Ontological concept in the ontology library; Recording user for each Ontological concept is set up user preference table, is determined user preference to this Ontological concept according to described degree of recognition to the degree of recognition of Ontological concept in user and system interaction process; This method comprises:
After receiving the Web service request, utilize the formalized description of described Web service gang form, in described Web service gang form storehouse, search Web service gang form with described Web service request coupling, if have the Web service gang form that mates fully with the Web service request, then described Web service gang form be back to the user; If do not have the Web service gang form that mates with the Web service request, then make up new Web service gang form:
Search each Web service with described Web service request coupling, in each Web service that is found out, choose the Web service of user preference degree maximum respectively by user's request, selected Web service is recommended to the user, all the other Web services are saved in the stack, if the Web service that user's acceptance is recommended, then as the Web service that makes up the Web service gang form, if the user does not accept the Web service recommended, then in described stack, carry out retrospective search, again with retrospective search to Web service recommend to the user; Determine whether each Web service need introduce new solicited message in the new Web service gang form of creating, and when having new solicited message, searches the Web service of satisfying this new solicited message, and recommends to the user; Retrospective search failure and new solicited message are searched when finishing, and will be made up the Web service gang form by the Web service that the user accepts, if the Web service of not accepted by the user, then the Web service gang form makes up failure;
If there is the Web service gang form that not exclusively mates with the Web service request, then based on the Web service gang form of described incomplete coupling, making up provides to the Web service gang form:
Do not search the Web service of coupling for there being request item in the Web service request that Web service matches, in each Web service that is found out, choose the Web service of user preference degree maximum respectively by user's request, selected Web service is recommended to the user, all the other Web services are saved in the stack, if the Web service that user's acceptance is recommended, then as the additional Web service of the Web service gang form of described incomplete coupling, if the user does not accept the Web service recommended, then in described stack, carry out retrospective search, again with retrospective search to Web service recommend to the user; Determine that whether described additional Web service need introduce new solicited message, when having new solicited message, searches the Web service of satisfying this new solicited message, and recommends to the user; Retrospective search failure and new solicited message are searched when finishing, and the Web service that will be accepted by the user makes up new Web service gang form as the additional Web clothes of the Web service gang form of described incomplete coupling;
Described Web service gang form is provided to the user, and for the Web service gang form of new structure, if accepted by the user, then the Web service gang form with described new structure is added into described Web service gang form storehouse.
2, method according to claim 1 is characterized in that, described Web service specifically describes and is: and WebService (WSInput, WSOutput, WSPrecondition, WSEffect), wherein, WebService is the title of Web service; WSInput is the input information set of service; WSOutput is the output information set of service; WSPrecondition is the precondition of service; WSEffect is the influence that service produces; Service request is described concrete structure: and RequestWS (RWSInput, RWSOutput), wherein, RequestWS is the title of Web service request; RWSInput is the input information set that the user provides; RWSOutput is the target information set of user's service request; Everyly in described Web service, the Web service request all describe by the Ontological concept in the ontology library.
3, method according to claim 1, it is characterized in that, described Web service gang form profile specifically describes: TemplateProfile (TemplateInput, TemplateOutput, TemplateCategory, Precondition, Effect), wherein, TemplateInput is the input information of described Web service gang form, corresponding more than one Ontological concept; TemplateOutput is the output information of described Web service gang form, also corresponding more than one Ontological concept; TemplateCategory represents the classified information of described Web service gang form; Precondition represents the precondition of described Web service gang form; Effect represents effect and the influence after described Web service gang form is carried out; The tabulation of described Web abstract service comprises more than one Web abstract service, described Web abstract service specifically describes: AbstractWS (AbstractWSInput, AbstractWSOutput, Precondition, Effect), wherein, AbstractWSInput is the input information of described Web abstract service, corresponding more than one Ontological concept; AbstractWSOutput is the output information of described Web abstract service, also corresponding more than one Ontological concept; AbstractWSCategory is the classified information of described Web abstract service; Precondition is the precondition of described Web abstract service; Effect is effect and the influence after described Web abstract service is carried out.
According to claim 2 or 3 described methods, it is characterized in that 4, the user preference degree of described Ontological concept is pressed following formula and determined:
Wherein, N=(Ontological concept is as the number of times of user's service request input information) * a%+ (Ontological concept is as the number of times of user's service request output information) * b%+ (number of times that number of times * 2-of being accepted by the user in recommendation process of Ontological concept is refused by the user) * c%; Wherein, the value of a is 10 to 20, and the value of b is 30 to 40, and the value of c is 50 to 60;
Matching degree between described Web service gang form and the user's Web service request RequestWS is pressed following formula and is determined:
Wherein, RWSInput
iMatching degree=the RWSInput of notion
iThe matching degree of the notion of mating most in the input information set of notion and Web composite services template; RWSOutput
jMatching degree=the RWSOutput of notion
jThe matching degree of the notion of mating most in the output information set of notion and Web composite services template.
5, according to claim 2 or 3 described methods, it is characterized in that, determine that the function of the user preference of Web service is:
f(n)＝g(n)+h(n)
Wherein,
The Input notion number of n for not provided by the user at present in the Web service, Input
iOntological concept in i the parameter in the Input information that expression n can not be provided by the user.
6, method according to claim 1 is characterized in that, described logical relation comprises order, parallel and selection at least.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008102256163A CN101383854B (en) | 2008-10-31 | 2008-10-31 | Web service providing method based on customer preference |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008102256163A CN101383854B (en) | 2008-10-31 | 2008-10-31 | Web service providing method based on customer preference |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101383854A true CN101383854A (en) | 2009-03-11 |
| CN101383854B CN101383854B (en) | 2012-01-25 |
Family
ID=40463473
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2008102256163A Expired - Fee Related CN101383854B (en) | 2008-10-31 | 2008-10-31 | Web service providing method based on customer preference |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101383854B (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102135998A (en) * | 2011-03-24 | 2011-07-27 | 浙江大学 | Method for interactively combining Web services based on precondition and post-condition |
| CN103207858A (en) * | 2012-01-11 | 2013-07-17 | 富士通株式会社 | Device and method for recommending Web service combination |
| CN103763302A (en) * | 2013-12-16 | 2014-04-30 | 东南大学 | Web service combination generating method |
| US8756323B2 (en) | 2010-11-26 | 2014-06-17 | International Business Machines Corporation | Semantic- and preference-based planning of cloud service templates |
| CN103986704A (en) * | 2014-05-09 | 2014-08-13 | 东南大学 | Connection-oriented network service construction method |
| CN104008109A (en) * | 2013-02-26 | 2014-08-27 | 南京邮电大学 | User interest based Web information push service system |
| TWI473028B (en) * | 2012-05-09 | 2015-02-11 | Univ Nat Pingtung Sci & Tech | Method for recommending a web service and computer product thereof |
| CN105786460A (en) * | 2014-12-24 | 2016-07-20 | 华为技术有限公司 | Software development method and device based on service composition |
| CN107169727A (en) * | 2017-04-20 | 2017-09-15 | 青岛科技大学 | Towards the cloud service combined method and platform of Making mold |
| CN110135747A (en) * | 2019-05-21 | 2019-08-16 | 成都派沃特科技股份有限公司 | Process customizing method neural network based |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR2906383B1 (en) * | 2006-09-21 | 2009-04-03 | Alcatel Sa | SEMANTIC WEB SERVICE REFERENTIAL AND METHOD USING THE REFERENTIAL |
-
2008
- 2008-10-31 CN CN2008102256163A patent/CN101383854B/en not_active Expired - Fee Related
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8756323B2 (en) | 2010-11-26 | 2014-06-17 | International Business Machines Corporation | Semantic- and preference-based planning of cloud service templates |
| US8762539B2 (en) | 2010-11-26 | 2014-06-24 | International Business Machines Corporation | Semantic- and preference-based planning of cloud service templates |
| CN102135998A (en) * | 2011-03-24 | 2011-07-27 | 浙江大学 | Method for interactively combining Web services based on precondition and post-condition |
| CN103207858B (en) * | 2012-01-11 | 2016-03-23 | 富士通株式会社 | For recommending the apparatus and method of Web service combination |
| CN103207858A (en) * | 2012-01-11 | 2013-07-17 | 富士通株式会社 | Device and method for recommending Web service combination |
| TWI473028B (en) * | 2012-05-09 | 2015-02-11 | Univ Nat Pingtung Sci & Tech | Method for recommending a web service and computer product thereof |
| CN104008109B (en) * | 2013-02-26 | 2017-11-14 | 南京邮电大学 | Web information Push Service system based on user interest |
| CN104008109A (en) * | 2013-02-26 | 2014-08-27 | 南京邮电大学 | User interest based Web information push service system |
| CN103763302B (en) * | 2013-12-16 | 2017-01-25 | 东南大学 | Web service combination generating method |
| CN103763302A (en) * | 2013-12-16 | 2014-04-30 | 东南大学 | Web service combination generating method |
| CN103986704A (en) * | 2014-05-09 | 2014-08-13 | 东南大学 | Connection-oriented network service construction method |
| CN105786460A (en) * | 2014-12-24 | 2016-07-20 | 华为技术有限公司 | Software development method and device based on service composition |
| CN105786460B (en) * | 2014-12-24 | 2019-02-26 | 华为技术有限公司 | A kind of software development methodology and device based on Services Composition |
| CN107169727A (en) * | 2017-04-20 | 2017-09-15 | 青岛科技大学 | Towards the cloud service combined method and platform of Making mold |
| CN107169727B (en) * | 2017-04-20 | 2024-03-29 | 青岛科技大学 | Cloud service combination method and system for mold manufacturing |
| CN110135747A (en) * | 2019-05-21 | 2019-08-16 | 成都派沃特科技股份有限公司 | Process customizing method neural network based |
| CN110135747B (en) * | 2019-05-21 | 2023-05-05 | 成都派沃特科技股份有限公司 | Flow customization method based on neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101383854B (en) | 2012-01-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101383854B (en) | Web service providing method based on customer preference | |
| Schneider et al. | A survey of the standard location-routing problem | |
| Green | Beyond value engineering: SMART value management for building projects | |
| CN102004767A (en) | Abstract service logic-based interactive semantic Web service dynamic combination method | |
| CN103838857B (en) | Automatic service combination system and method based on semantics | |
| Bagheri et al. | Solving the fully fuzzy multi-objective transportation problem based on the common set of weights in DEA | |
| Liu et al. | Customer-driven product design and evaluation method for collaborative design environments | |
| Rostami et al. | A lagrangian relaxation algorithm for facility location of resource-constrained decentralized multi-project scheduling problems | |
| Morabit et al. | Machine-learning–based arc selection for constrained shortest path problems in column generation | |
| CN112052322B (en) | Intelligent robot conversation strategy generation method based on particle calculation | |
| Herazo-Padilla et al. | Simulation-optimization approach for the stochastic location-routing problem | |
| Djamila et al. | A temporal distributed group decision support system based on multi-criteria analysis | |
| Frías et al. | Hybrid Algorithms for energy minimizing vehicle routing problem: integrating clusterization and ant colony optimization | |
| Nguyen et al. | A genetic programming approach for evolving variable selectors in constraint programming | |
| Murray-Rust et al. | Softening electronic institutions to support natural interaction | |
| Omari et al. | Coupling multi-criteria analysis and machine learning for agent based group decision support: Spatial localization | |
| Crispim et al. | Multiple criteria partner selection in virtual enterprises | |
| Soroor et al. | An advanced adoption model and an algorithm of evaluation agents in automated supplier ranking | |
| Mourad et al. | Ontology-Based Reasoning System for Logistics Applications Deployment | |
| Fasanghari et al. | The fuzzy evaluation of the ict projects in strategic environment (case study: Iran telecommunication research center) | |
| Du et al. | An Explanation Oriented Dialogue Approach for Solving Wicked Planning Problems. | |
| Scholz-Reiter et al. | Modeling of orders in autonomously controlled logistic systems | |
| CN103973799B (en) | A kind of resource environment data decision support platform | |
| Bakhshi et al. | A fuzzy-based approach for selecting the optimal composition of services according to user preferences | |
| Zhou et al. | A novel Bi-level programming model for cloud logistics resources allocation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20120125 Termination date: 20141031 |
|
| EXPY | Termination of patent right or utility model |