WO2012116512A1 - Service discovery method and system based on multiple features matching - Google Patents
Service discovery method and system based on multiple features matching Download PDFInfo
- Publication number
- WO2012116512A1 WO2012116512A1 PCT/CN2011/072647 CN2011072647W WO2012116512A1 WO 2012116512 A1 WO2012116512 A1 WO 2012116512A1 CN 2011072647 W CN2011072647 W CN 2011072647W WO 2012116512 A1 WO2012116512 A1 WO 2012116512A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- request message
- query request
- service
- functional
- matrix
- Prior art date
Links
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
Definitions
- the present invention relates to the field of computer technology, and more particularly to a service discovery method and system based on multi-feature matching. Background technique
- the SOC environment is developing rapidly in both directions.
- the network access mode is developed by a single access method combining fixed Internet and ordinary computers to multiple access modes and multiple terminal devices.
- the scale and scope of large-scale distributed applications are expanding, and the demand for services across multiple autonomous areas is exploding.
- user needs are constantly changing, service providers are constantly improving their services, and new networks will appear at any time. Service, there will be a service exit at any time.
- the semantic service discovery method is mainly based on direct reasoning.
- the service request matches the candidate service, it often involves semantic relationship judgment between multiple parameters across different ontology domains, and the semantic relationship.
- Judgment involves a large amount of ontology loading, classification, and reasoning, so time efficiency is very low.
- the object of the present invention is to provide a service discovery method and system based on multi-feature matching, which improves the time efficiency of obtaining a service.
- the embodiment of the invention provides a service discovery method based on multi-feature matching, which includes: Acquiring each of the candidate service sets associated with the query request message according to the query request message acquisition request input functional matrix set and the request output functional matrix according to the request input functional matrix set and the request output functional matrix set Similarity between services;
- the embodiment of the present invention provides a service discovery system based on multi-feature matching.
- the service discovery system based on multi-feature matching can implement the service discovery method based on multi-feature matching of the above technical solution, where the system includes:
- a first acquiring module configured to acquire a request input functional matrix set and a request output functional matrix set according to the query request message
- a calculation module configured to calculate, according to the request input functional matrix set and the request output functional matrix set, a similarity between each of the candidate service sets associated with the query request message;
- a second obtaining module configured to obtain, according to the similarity, a service list to be queried by the query request message.
- the multi-feature matching based service discovery method and system provided by the present invention calculates the similarity between each service in the candidate service set associated with the query request message according to the request input functional matrix set and the request output functional matrix set. And obtaining the service list to be queried according to the similarity degree, and avoiding the reasoning of the semantics of the service and calculating the similarity of the service when the service is acquired, thereby improving the matching efficiency between the service and the request query message, thereby Increased time efficiency in getting services.
- FIG. 1 is a schematic flow chart of an embodiment of a service discovery method based on multi-feature matching according to the present invention
- FIG. 2 is a schematic flow chart of still another embodiment of a service discovery method based on multi-feature matching according to the present invention
- FIG. 3 is a schematic structural diagram of an embodiment of a service discovery system based on multi-feature matching according to the present invention.
- FIG. 4 is a schematic structural diagram of still another embodiment of a service discovery system based on multi-feature matching according to the present invention. detailed description
- FIG. 1 is a schematic flow chart of an embodiment of a service discovery method based on multi-feature matching according to the present invention. As shown in FIG. 1, the embodiment of the present invention includes the following steps:
- Step 101 Acquire a request input functional matrix set and a request output functional matrix set according to the query request message
- Step 102 Enter a functional matrix set and a request output functional matrix set according to the request, and calculate a similarity between each service in the candidate service set associated with the query request message.
- Step 103 Acquire, according to the similarity, a service list to be queried by the query request message.
- the multi-feature matching based service discovery method calculates the similarity between each service in the candidate service set associated with the query request message according to the request input functional matrix set and the request output functional matrix set. And get the query request message based on the similarity
- the service list to be queried avoids the reasoning of the semantics of the service and the similarity of the service when the service is obtained, thereby improving the matching efficiency between the service and the request query message, and further improving the time efficiency of obtaining the service.
- FIG. 2 is a schematic flow chart of still another embodiment of a service discovery method based on multi-feature matching according to the present invention. As shown in FIG. 2, the embodiment of the present invention includes the following steps:
- Step 201 Perform semantic information extraction on the publishing document corresponding to the query request message, and obtain the number of all the bodies of the query request message, the input parameters of the query request message, and the semantic concepts of all input parameters in the plurality of ontology referenced by the query request message. Relationship;
- Step 202 Acquire a request input functional matrix set according to the relationship between the number of all the ontology, the input parameter of the query request message, the output parameter of the query request message, and the semantic concepts of all the input parameters in the plurality of ontology referenced by the query request message. ;
- Step 203 Acquire a request output function matrix set according to the relationship between the number of all the ontology, the output parameter of the query request message, and the semantic concepts of all output parameters in the plurality of ontology referenced by the query request message.
- Step 204 Filter a plurality of services stored in the database according to the query request message.
- Step 205 Perform, according to the type information of the multiple services, the number of input and output of the multiple services, and the ontology information referenced by the service. Filtering to obtain a candidate service set associated with the query request message;
- Step 206 Calculate a function similarity matrix of each service in the candidate service set and the query request message according to a service functional matching method
- Step 207 Calculate, according to the service non-functional matching method, each service in the candidate service set and a non-functional similarity matrix of the query request message.
- Step 208 Acquire a text similarity matrix according to a text feature vector of each service in the candidate service set and a text feature vector of the query request message.
- Step 209 Acquire, according to the functional similarity matrix, the non-functional similarity matrix, and the text similarity matrix, a similarity between the query request message and each service in the candidate service.
- Step 210 Acquire, according to the similarity, a service list to be queried by the query request message.
- the request semantic information pre-processing mechanism extracts the request metadata information and the request parameter semantic information while the query request is issued, performs consistency check on the request semantics, and saves the pre-processed result;
- the embodiment presets a semantic information extraction template (SRT) for service request and a service publishing template (SPT) for semantic information extraction; wherein, according to the service semantic metadata information preset in the service publishing template (SPT), Querying the semantic information in the request message R, since the SRT is the basis for extracting the semantic information of the query request message, by parsing the SRT, the semantic information in the query request message can be obtained, and the service functional matrix in the prior art, due to the query
- the semantic information preprocessing stage in the request message does not require semantic reasoning, thus making the subsequent request input functional matrix set and output functional matrix set generation process relatively simple.
- the request input functional matrix set FRi may specifically be: FRi ⁇ FR,
- the request input functional matrix set FRi is composed of k matrices, k is the number of all reference ontology of the query request message R, and the matrix FRij represents the input parameter of the query request message R and the query request message R reference body of the j-th input parameters for all relationships semantic concept; FRij number of rows of the matrix of the input parameters for the number of columns in the number of R m, FRij matrix body ont k is the number of concepts cn k; specifically, in accordance with formula
- FRoj.Vpq is the value of the element in the jth matrix of the pth row in the output matrix set of the query request message R;
- FRoj.Vpq represents the R output on p semantic concept and the ontology in ontotj
- the exact degree of matching of the q concepts ontj .conceptq requires that the elements in the output functional matrix have only two values, 1 and 0.
- the candidate service set may be filtered based on the service category; then, the candidate service set is filtered based on the service input and output number, and finally the candidate service set is filtered based on the service referenced ontology; specifically, the service category is The description of the domain to which the service function belongs is selectively given by the service provider when the service is published.
- the candidate service set is selected according to the category.
- Each service that includes a service category constraint is traversed, and when the service category is not included in the requested service category, the service is filtered out; when the candidate service set is filtered based on the number of service input and output, the premise hypothesis service output is introduced.
- the service cannot provide the complete information required for the query request message, and the service and the query request message cannot be matched; when the number of service input parameters is greater than the input of the query request message.
- the number of numbers it is considered that the query request message cannot provide all the inputs required to satisfy the service, and the service also fails to match the query request message; based on the above description, the number of input parameters in the candidate service set is greater than the input parameter of the query request message.
- the services of the trees are filtered out; the services of the output parameters of the database are smaller than the number of output parameters of the query request message; in addition, due to any domain ontology, the description of the concept in the field is complete, different ontology descriptions Collar
- the domain is also different, so if the service does not reference the ontology referenced in the request, then the service cannot provide the same functional parameters as the functional parameters required by the query request message, and there is no possibility between the service and the query request message. Similarity relationship, so the services in the service candidate set that do not reference the referenced ontology can be filtered out by the above process.
- the service functional matching method is specifically: inputting a functional matrix set for a given service
- iResult consists of k matrices, k is the maximum of the number of all referenced ontology of the service S and query request message R input parameters; the above k ontology constitutes the ontology set of the service matching phase For any ontology ontj, if the output parameter of service S or query request message R is not referenced ⁇ , then the corresponding service functional matrix set or the jth matrix element of the request functional matrix set is replaced with all zeros; iResult The mth element is Fi m ⁇ FRi ⁇ which represents the product of the input functional matrix Fi m of the service S and the input functional matrix FRi donne ⁇ of the query request message R.
- the service non-functional matching method is specifically: accepting any service S and the query request message R in the service service candidate set, and calculating and returning the non-functional attribute matching similarity between the service S and the query request message R;
- the attribute qo of each quality of service (QoS) defined in the message R is judged whether it also defines the same attribute qo in the service S, and if so, the value range and the query request message of the attribute in the service S are calculated.
- QoS quality of service
- the overlap of the attribute value range occupies the ratio of the value range of the attribute in the query request message R, and the ratio is the similarity between the service S and the query request message R on the qosj attribute; finally, the similarity of all QoS attributes is calculated.
- the average value is used to calculate the non-functional matching similarity between the service S and the query request message R.
- the functional similarity matrix, the non-functional similarity matrix, and the text similarity matrix are merged, and the services are sorted according to the final similarity, according to the functional similarity ms,
- the multi-feature matching based service discovery method calculates the similarity between each service in the candidate service set associated with the query request message according to the request input functional matrix set and the request output functional matrix set. And obtaining the service list to be queried according to the similarity degree, and avoiding the reasoning of the semantics of the service and the similarity of the service when the service is acquired, thereby improving the matching efficiency between the service and the request query message, The time efficiency of obtaining services is further improved.
- FIG. 3 is a schematic structural diagram of an embodiment of a service discovery system based on multi-feature matching according to the present invention. As shown in FIG. 3, the embodiment of the present invention includes: a first acquisition module 31, a calculation module 32, and a second acquisition module 33;
- the first obtaining module 31 acquires a request input functional matrix set and a request output functional matrix set according to the query request message; the calculating module 32 calculates and the query according to the request input functional matrix set and the request output functional matrix set.
- the similarity between each of the candidate service sets associated with the request message; the second obtaining module 33 acquires the service list to be queried by the query request message according to the similarity.
- the multi-feature matching based service discovery system calculates, by the calculation module 32, each service in the candidate service set associated with the query request message according to the request input functional matrix set and the request output functional matrix set.
- the second acquisition module 33 obtains the service list to be queried according to the similarity degree according to the similarity degree, and improves the service by avoiding the reasoning of the semantics of the service and calculating the similarity of the service when acquiring the service.
- the efficiency of matching with the request query message further improves the time efficiency of obtaining the service.
- FIG. 4 is a schematic structural diagram of still another embodiment of a service discovery system based on multi-feature matching according to the present invention.
- the embodiment of the present invention includes: a first obtaining module 41, a calculating module 42, and a second Obtaining module 43;
- the first obtaining module 41 acquires a request input functional matrix set and a request output functional matrix set according to the query request message; the calculating module 42 calculates and the query according to the request input functional matrix set and the request output functional matrix set.
- the first obtaining module 41 may further include: an extracting unit 411, a first obtaining unit 412, and a second obtaining unit 413; wherein, the extracting unit 411 performs semantic information extraction on the published document corresponding to the query request message, to obtain The relationship between the number of all the ontology of the query request message, the input parameter of the query request message, and the semantic concept of all the input parameters in the plurality of ontology referenced by the query request message; the first obtaining unit 412 is configured according to the number of all the ontology The input parameter of the query request message, the output parameter of the query request message, and the relationship between all the input parameter semantic concepts in the plurality of ontology referenced by the query request message acquire the request input functional matrix set; the second obtaining unit 413 And obtaining a request output functional matrix set according to the relationship between the number of all the ontology, the output parameter of the query request message, and the semantic concepts of all output parameters in the plurality of ontology referenced by the query request message.
- the calculation module 42 may further include: a first filtering unit 421, a second filtering unit 422, and a first calculating unit 423; wherein, the first filtering unit 421 searches for multiple services stored in the database according to the query request message. Performing filtering; the second filtering unit 422 filters the service according to the type information of the multiple services, the number of input and output of the multiple services, and the ontology information used by the service, to obtain the association with the query request message. a candidate service set; the first computing unit 423 calculates a similarity of each of the candidate service sets to the query request message.
- the first calculating unit 423 may further include: a first calculating subunit, a second calculating subunit, a first acquiring subunit, and a second acquiring subunit; wherein the first calculating subunit is calculated according to a service functional matching method Each of the candidate service sets and the query request cancellation a functional similarity matrix; the second computing subunit calculates a non-functional similarity matrix of each of the candidate service sets and the query request message according to a service non-functional matching method; a text feature vector of each service in the candidate service set and a text feature vector of the query request message to obtain a text similarity matrix; the second obtaining subunit is based on the functional similarity matrix, the non-functional similarity matrix, and the text similarity The degree matrix obtains the similarity between the query request message and each of the candidate services.
- the second obtaining subunit may further include: a multiplication subunit and a calculation subunit; wherein, the multiplication subunit is configured to perform matrix matching on the functional similarity matrix and the non-functional similarity matrix to generate an input matching result matrix; a subunit, configured to calculate a number of all non-zero elements in the matching result matrix, the ratio of the number of all non-zero elements to the number of all input parameters and output parameters as the service and the query request message The similarity of the match.
- the multi-feature matching based service discovery method calculates, by the calculation module 42, each service in the candidate service set associated with the query request message according to the request input functional matrix set and the request output functional matrix set.
- the second obtaining module 43 obtains the service list to be queried according to the similarity degree according to the similarity degree, and improves the service because the semantics of the service are avoided and the similarity of the service is calculated when the service is acquired.
- the efficiency of matching with the request query message further improves the time efficiency of obtaining the service.
- the foregoing program may be stored in a computer readable storage medium, and when executed, the program includes The foregoing steps of the method embodiment; and the foregoing storage medium includes: a medium that can store program codes, such as a ROM, a RAM, a magnetic disk, or an optical disk.
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
A service discovery method and system based on multiple features matching are disclosed. The method comprises: acquiring a functional matrix set of request input and a functional matrix set of request output according to a query request message; computing similarity between every service of candidate service set associating with the query request message according to the functional matrix set of request input and the functional matrix set of request output; acquiring a service list to be inquired by the query request message according to the similarity. The service discovery method and system of the present invention avoid reasoning to service semanteme and computing service similarity when discovering service, thereby improve matching efficiency of the service and request message query, and acquirement service efficiency.
Description
基于多特征匹配的服务发现方法及*** 技术领域 Service discovery method and system based on multi-feature matching
本发明涉及计算机技术领域， 尤其是一种基于多特征匹配的服务发现方 法及***。 背景技术 The present invention relates to the field of computer technology, and more particularly to a service discovery method and system based on multi-feature matching. Background technique
互联网的迅速发展使得现有网络成为一个巨大的异构平台， 大量的软件 服务部署于该异构平台上。 用户如何方便地使用这些服务， 服务之间如何进 行灵活的互操作成为学术界与工业界共同关注的重要问题， 由此产生了面向 月良务的计算（Service-Oriented Computing, 简称： SOC )。 The rapid development of the Internet has made the existing network a huge heterogeneous platform, and a large number of software services are deployed on the heterogeneous platform. How users can easily use these services, and how flexible interoperability between services has become an important issue for both academic and industrial circles, resulting in Service-Oriented Computing (SOC).
目前， SOC环境正向两个方向飞速发展， 一方面， 网络访问方式由过去 单一的固定互联网与普通计算机结合的访问方式向着多种接入、 多种终端设 备参与的访问方式发展； 另一方面， 大型分布式应用的规模和范围不断扩大， 跨越多个自治领域的服务需求激增； 在实际的服务环境中， 用户需求不断变 化， 服务提供者也不断的改进其服务， 网络中随时会出现新服务， 也随时会 有服务退出。 At present, the SOC environment is developing rapidly in both directions. On the one hand, the network access mode is developed by a single access method combining fixed Internet and ordinary computers to multiple access modes and multiple terminal devices. The scale and scope of large-scale distributed applications are expanding, and the demand for services across multiple autonomous areas is exploding. In the actual service environment, user needs are constantly changing, service providers are constantly improving their services, and new networks will appear at any time. Service, there will be a service exit at any time.
在现有的 Web服务解决方案中,语义服务发现方法主要基于直接推理， 在服务请求与候选服务进行服务匹配时， 往往涉及跨越不同本体领域的多个 参数之间的语义关系判断， 而语义关系判断涉及大量本体装载、 分类和推理， 因此时间效率十分低下。 发明内容 In the existing Web service solution, the semantic service discovery method is mainly based on direct reasoning. When the service request matches the candidate service, it often involves semantic relationship judgment between multiple parameters across different ontology domains, and the semantic relationship. Judgment involves a large amount of ontology loading, classification, and reasoning, so time efficiency is very low. Summary of the invention
本发明的目的在于提供一种基于多特征匹配的服务发现方法及***， 提 高获取服务的时间效率。 The object of the present invention is to provide a service discovery method and system based on multi-feature matching, which improves the time efficiency of obtaining a service.
本发明实施例提供一种基于多特征匹配的服务发现方法， 包括：
根据查询请求消息获取请求输入功能性矩阵集合和请求输出功能性矩阵 根据所述请求输入功能性矩阵集合和请求输出功能性矩阵集合计算与所 述查询请求消息相关联的候选服务集合中的每一个服务之间的相似度； The embodiment of the invention provides a service discovery method based on multi-feature matching, which includes: Acquiring each of the candidate service sets associated with the query request message according to the query request message acquisition request input functional matrix set and the request output functional matrix according to the request input functional matrix set and the request output functional matrix set Similarity between services;
根据所述相似度获取所述查询请求消息待查询的服务列表。 Obtaining, according to the similarity, a service list to be queried by the query request message.
本发明实施例提供一种基于多特征匹配的服务发现***， 该基于多特征 匹配的服务发现***能够实现上述技术方案的基于多特征匹配的服务发现方 法， 其中， 该***包括： The embodiment of the present invention provides a service discovery system based on multi-feature matching. The service discovery system based on multi-feature matching can implement the service discovery method based on multi-feature matching of the above technical solution, where the system includes:
第一获取模块， 用于根据查询请求消息获取请求输入功能性矩阵集合和 请求输出功能性矩阵集合； a first acquiring module, configured to acquire a request input functional matrix set and a request output functional matrix set according to the query request message;
计算模块， 用于根据所述请求输入功能性矩阵集合和请求输出功能性矩 阵集合计算与所述查询请求消息相关联的候选服务集合中的每一个服务之间 的相似度； a calculation module, configured to calculate, according to the request input functional matrix set and the request output functional matrix set, a similarity between each of the candidate service sets associated with the query request message;
第二获取模块， 用于根据所述相似度获取所述查询请求消息待查询的服 务列表。 And a second obtaining module, configured to obtain, according to the similarity, a service list to be queried by the query request message.
本发明提供的基于多特征匹配的服务发现方法及***， 根据请求输入功 能性矩阵集合和请求输出功能性矩阵集合计算与查询请求消息相关联的候选 服务集合中的每一个服务之间的相似度， 并根据相似度获取与查询请求消息 待查询的服务列表， 由于在获取服务时避免了对服务的语义进行推理以及对 服务的相似度进行计算，从而提高服务与请求查询消息的匹配效率，从而提高 了获取服务的时间效率。 附图说明 The multi-feature matching based service discovery method and system provided by the present invention calculates the similarity between each service in the candidate service set associated with the query request message according to the request input functional matrix set and the request output functional matrix set. And obtaining the service list to be queried according to the similarity degree, and avoiding the reasoning of the semantics of the service and calculating the similarity of the service when the service is acquired, thereby improving the matching efficiency between the service and the request query message, thereby Increased time efficiency in getting services. DRAWINGS
为了更清楚地说明本发明实施例或现有技术中的技术方案， 下面将对实 施例或现有技术描述中所需要使用的附图作简单地介绍， 显而易见地， 下面 描述中的附图仅仅是本发明的一些实施例， 对于本领域普通技术人员来讲，
在不付出创造性劳动的前提下， 还可以根据这些附图获得其他的附图。 In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described below. Obviously, the drawings in the following description are only It is some embodiments of the present invention, and those of ordinary skill in the art, Other drawings may also be obtained from these drawings without paying for creative labor.
图 1 为本发明基于多特征匹配的服务发现方法一个实施例的流程示意 图； 1 is a schematic flow chart of an embodiment of a service discovery method based on multi-feature matching according to the present invention;
图 2为本发明基于多特征匹配的服务发现方法又一个实施例的流程示意 图； 2 is a schematic flow chart of still another embodiment of a service discovery method based on multi-feature matching according to the present invention;
图 3 为本发明基于多特征匹配的服务发现***一个实施例的结构示意 图； 3 is a schematic structural diagram of an embodiment of a service discovery system based on multi-feature matching according to the present invention;
图 4为本发明基于多特征匹配的服务发现***又一个实施例的结构示意 图。 具体实施方式 4 is a schematic structural diagram of still another embodiment of a service discovery system based on multi-feature matching according to the present invention. detailed description
下面将结合本发明实施例中的附图， 对本发明实施例中的技术方案进行 清楚、 完整地描述， 显然， 所描述的实施例仅仅是本发明一部分实施例， 而 不是全部的实施例。 基于本发明中的实施例， 本领域普通技术人员在没有做 出创造性劳动前提下所获得的所有其他实施例， 都属于本发明保护的范围。 The technical solutions in the embodiments of the present invention are clearly and completely described in the following with reference to the accompanying drawings in the embodiments of the present invention. It is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts are within the scope of the present invention.
图 1 为本发明基于多特征匹配的服务发现方法一个实施例的流程示意 图， 如图 1所示， 本发明实施例包括如下步骤： FIG. 1 is a schematic flow chart of an embodiment of a service discovery method based on multi-feature matching according to the present invention. As shown in FIG. 1, the embodiment of the present invention includes the following steps:
步骤 101、 根据查询请求消息获取请求输入功能性矩阵集合和请求输出 功能性矩阵集合； Step 101: Acquire a request input functional matrix set and a request output functional matrix set according to the query request message;
步骤 102、 根据该请求输入功能性矩阵集合和请求输出功能性矩阵集合 计算与该查询请求消息相关联的候选服务集合中的每一个服务之间的相似 度； Step 102: Enter a functional matrix set and a request output functional matrix set according to the request, and calculate a similarity between each service in the candidate service set associated with the query request message.
步骤 103、 根据该相似度获取该查询请求消息待查询的服务列表。 Step 103: Acquire, according to the similarity, a service list to be queried by the query request message.
本发明实施例提供的基于多特征匹配的服务发现方法， 根据请求输入功 能性矩阵集合和请求输出功能性矩阵集合计算与查询请求消息相关联的候选 服务集合中的每一个服务之间的相似度， 并根据相似度获取与查询请求消息
待查询的服务列表， 由于在获取服务时避免了对服务的语义进行推理以及对 服务的相似度进行计算， 从而提高了服务与请求查询消息的匹配效率， 进一 步提高了获取服务的时间效率。 The multi-feature matching based service discovery method provided by the embodiment of the present invention calculates the similarity between each service in the candidate service set associated with the query request message according to the request input functional matrix set and the request output functional matrix set. And get the query request message based on the similarity The service list to be queried avoids the reasoning of the semantics of the service and the similarity of the service when the service is obtained, thereby improving the matching efficiency between the service and the request query message, and further improving the time efficiency of obtaining the service.
图 2为本发明基于多特征匹配的服务发现方法又一个实施例的流程示意 图， 如图 2所示， 本发明实施例包括如下步骤： 2 is a schematic flow chart of still another embodiment of a service discovery method based on multi-feature matching according to the present invention. As shown in FIG. 2, the embodiment of the present invention includes the following steps:
步骤 201、 对查询请求消息相对应的发布文档进行语义信息抽取， 得到 该查询请求消息所有本体的数目、 该查询请求消息的输入参数与该查询请求 消息引用的多个本体中所有输入参数语义概念的关系； Step 201: Perform semantic information extraction on the publishing document corresponding to the query request message, and obtain the number of all the bodies of the query request message, the input parameters of the query request message, and the semantic concepts of all input parameters in the plurality of ontology referenced by the query request message. Relationship;
步骤 202、 根据该所有本体的数目、 该查询请求消息的输入参数、 该查 询请求消息的输出参数与该查询请求消息引用的多个本体中所有输入参数语 义概念的关系获取请求输入功能性矩阵集合； Step 202: Acquire a request input functional matrix set according to the relationship between the number of all the ontology, the input parameter of the query request message, the output parameter of the query request message, and the semantic concepts of all the input parameters in the plurality of ontology referenced by the query request message. ;
步骤 203、 根据该所有本体的数目、 该查询请求消息的输出参数与该查 询请求消息引用的多个本体中所有输出参数语义概念的关系获取请求输出功 能性矩阵集合； Step 203: Acquire a request output function matrix set according to the relationship between the number of all the ontology, the output parameter of the query request message, and the semantic concepts of all output parameters in the plurality of ontology referenced by the query request message.
步骤 204、 根据该查询请求消息对保存在数据库中的多个服务进行过滤； 步骤 205、 根据该多个服务的类型信息、 该多个服务的输入输出数量以 及服务引用到的本体信息对服务进行过滤， 得到与该查询请求消息相关联的 候选服务集合； Step 204: Filter a plurality of services stored in the database according to the query request message. Step 205: Perform, according to the type information of the multiple services, the number of input and output of the multiple services, and the ontology information referenced by the service. Filtering to obtain a candidate service set associated with the query request message;
步骤 206、 根据服务功能性匹配方法计算该候选服务集合中的每一个服 务和所述查询请求消息的功能相似度矩阵； Step 206: Calculate a function similarity matrix of each service in the candidate service set and the query request message according to a service functional matching method;
步骤 207、 根据服务非功能性匹配方法计算该候选服务集合中的每一个 服务和该查询请求消息的非功能相似度矩阵； Step 207: Calculate, according to the service non-functional matching method, each service in the candidate service set and a non-functional similarity matrix of the query request message.
步骤 208、 根据该候选服务集合中的每一个服务的文本特征向量和该查 询请求消息的文本特征向量获取文本相似度矩阵； Step 208: Acquire a text similarity matrix according to a text feature vector of each service in the candidate service set and a text feature vector of the query request message.
步骤 209、 根据该功能相似度矩阵、 非功能相似度矩阵和文本相似度矩 阵获取该查询请求消息与该候选服务中的每一个服务的相似度；
步骤 210、 根据该相似度获取该查询请求消息待查询的服务列表。 Step 209: Acquire, according to the functional similarity matrix, the non-functional similarity matrix, and the text similarity matrix, a similarity between the query request message and each service in the candidate service. Step 210: Acquire, according to the similarity, a service list to be queried by the query request message.
上述步骤 201中， 当查询请求消息 R发布时， 请求语义信息预处理机制 在查询请求发布同时提取请求元数据信息及请求参数语义信息， 对请求语义 做一致性检查， 保存预处理结果； 本发明实施例预设了用于服务请求的语义 信息抽取模板 ( SRT ) 以及用于语义信息抽取的服务发布模板 ( SPT ); 其中， 根据服务发布模板 ( SPT )中预设的服务语义元数据信息抽取查询请求消息 R 中的语义信息， 由于 SRT是对查询请求消息进行语义信息抽取的基础， 通过 解析 SRT, 能够获得查询请求消息中的语义信息， 与现有技术中的服务功能 性矩阵， 由于查询请求消息中的语义信息预处理阶段不需要语义推理， 因此 使得后续的请求输入功能性矩阵集合与输出功能性矩阵集合生成过程相对简 单。 In the foregoing step 201, when the query request message R is issued, the request semantic information pre-processing mechanism extracts the request metadata information and the request parameter semantic information while the query request is issued, performs consistency check on the request semantics, and saves the pre-processed result; The embodiment presets a semantic information extraction template (SRT) for service request and a service publishing template (SPT) for semantic information extraction; wherein, according to the service semantic metadata information preset in the service publishing template (SPT), Querying the semantic information in the request message R, since the SRT is the basis for extracting the semantic information of the query request message, by parsing the SRT, the semantic information in the query request message can be obtained, and the service functional matrix in the prior art, due to the query The semantic information preprocessing stage in the request message does not require semantic reasoning, thus making the subsequent request input functional matrix set and output functional matrix set generation process relatively simple.
上述步骤 202中， 请求输入功能性矩阵集合 FRi具体可以为： FRi^FR , In the foregoing step 202, the request input functional matrix set FRi may specifically be: FRi^FR,
FRi2...FRiJ , 其中， 请求输入功能性矩阵集合 FRi由 k个矩阵组成， k为查 询请求消息 R所有引用本体的数目，矩阵 FRij表示查询请求消息 R的输入参 数与查询请求消息 R引用的第 j个本体中所有输入参数语义概念的关系； 矩 阵 FRij的行数为 R的输入参数中个数 m, 矩阵 FRij的列数为本体 ontk的概念 个 数 cnk ； 具 体 地 ， 可 以 按 照 公 式 FRi 2 ... FRiJ , wherein the request input functional matrix set FRi is composed of k matrices, k is the number of all reference ontology of the query request message R, and the matrix FRij represents the input parameter of the query request message R and the query request message R reference body of the j-th input parameters for all relationships semantic concept; FRij number of rows of the matrix of the input parameters for the number of columns in the number of R m, FRij matrix body ont k is the number of concepts cn k; specifically, in accordance with formula
^卜 , , , . , , , ， 、 ，— FRirvp = 到清求输入功能性矩阵
^卜, , , . , , , , , , — FRi r v p = to clear input functional matrix
集合 Fri, 其中， FRij.Vpq为查询请求消息 R输入矩阵集合中第 j个矩阵中位于 第 p 行 第 q 列 的 元 素 的 值 ； 基 于 公 式 fl,若 in = out . , concept ( l≤p≤m, l≤q≤cn .) . 、 + , , .The set Fri, where FRij.Vpq is the value of the element located in the qth column of the pth row in the jth matrix in the matrix of the query request message R; based on the formula fl, if in = out . , concept ( l ≤ p ≤ m, l≤q≤cn .) . , + , , .
= ， J ? FRij.Vpq代表 R输入 inp语= , J ? FRij. Vpq stands for R input in p
1 [ 0,其它 义概念与本体 ontj中的第 q个概念 ontj.conceptq的精确匹配程度，故请求输入 功能性矩阵中的元素只有 1和 0两个值。 1 [ 0, the exact meaning of the other semantic concept and the qth concept ontj.conceptq in the ontology ontj, so the elements in the input functional matrix are only 1 and 0.
上述步骤 203中， 查询请求消息 R的输出功能性矩阵集合 FRo具体可以
为： FRo^FRi , FRo2-FRok} , 其中， 输出功能性矩阵集合 FRo由 k个矩阵 组成， k为查询请求消息 R所有引用本体的数目； 矩阵 FRoj表示查询请求消 息 R的输出参数与查询请求消息 R引用的第 j个本体中所有输出参数语义概 念的关系； 矩阵 FRoj的行数为查询请求消息 R的输入参数个数 n, 矩阵 FRoj 的列数为本体 ontk 的概念个数 cnk； 具体地， 可以按照公式 il,¾:on >,= ont .. concept„( 1≤ p≤ m, 1≤ q≤ cn .) ^ In the foregoing step 203, the output functional matrix set FRo of the query request message R may be specifically Is: FRo ^ FRi, FRo 2 -FRo k}, wherein the output matrix FRo set of k matrices, k is the number of all query request message R references body; FRoj matrix represents an output parameter query request message and R Query the relationship between the semantic concepts of all output parameters in the jth ontology referenced by the request message R; the number of rows of the matrix FRoj is the number n of input parameters of the query request message R, and the number of columns of the matrix FRoj is the conceptual number of the ontot k of the ontology k ; Specifically, according to the formula il, 3⁄4 : on > , = ont .. concept„( 1 ≤ p ≤ m, 1 ≤ q ≤ cn .) ^
FRorvm = \ p 3 q 获取到输出功能性矩阵集合 FRo r v m = \ p 3 q Get the output functional matrix set
[0,其它 [0, other
Fro, 其中， FRoj.Vpq为查询请求消息 R的输出矩阵集合中第 j个矩阵中位于 第 p行第 q列的元素的值； FRoj.Vpq代表 R输出 onp语义概念与本体 ontj中的 第 q个概念 ontj .conceptq的精确匹配程度故请求输出功能性矩阵中的元素只有 1和 0两个值。 Fro, where FRoj.Vpq is the value of the element in the jth matrix of the pth row in the output matrix set of the query request message R; FRoj.Vpq represents the R output on p semantic concept and the ontology in ontotj The exact degree of matching of the q concepts ontj .conceptq requires that the elements in the output functional matrix have only two values, 1 and 0.
上述步骤 205中， 可以先基于服务类别对候选服务集合进行过滤； 然后 基于服务输入输出数目对候选服务集合进行过滤， 最后基于服务引用的本体 对候选服务集合进行过滤； 具体地， 服务类别是对服务功能所属领域的描述， 是由服务提供者发布服务时是选择性给出的，在基于服务类别进行过滤阶段， 如果服务请求者提供了服务类别约束， 则按照该类别对候选服务集合中的每 一个包含服务类别约束的服务进行遍历， 当存在服务类别不被请求的服务类 别所包含时， 将该服务过滤掉； 在基于服务输入输出数目对候选服务集合进 行过滤时， 引入前提假设服务输出参数数目少于查询请求消息要求的输出参 数数目时， 认为服务不能提供查询请求消息所需的完整信息， 则导致服务与 查询请求消息无法匹配； 当服务输入参数的数目大于查询请求消息的输入参 数的数目时， 认为查询请求消息不能提供满足服务所要求的全部输入， 则同 样导致服务与查询请求消息无法匹配； 基于上述描述， 将候选服务集合中输 入参数的数目大于查询请求消息的输入参数的树木的服务过滤掉； 将数据库 中的输出参数的树木小于查询请求消息的输出参数的数目的服务过滤掉； 此 外， 由于任一领域本体， 对本领域概念的描述是完备的， 不同本体描述的领
域也是不同的， 因此若服务没有引用请求中所引用的本体， 则服务不可能提 供与查询请求消息要求的功能参数相同领域的功能参数， 则该服务与查询请 求消息之间也不可能存在任何相似关系， 因此通过上述过程可以将服务候选 集合中没有引用所引用的本体的服务都过滤掉。 In the foregoing step 205, the candidate service set may be filtered based on the service category; then, the candidate service set is filtered based on the service input and output number, and finally the candidate service set is filtered based on the service referenced ontology; specifically, the service category is The description of the domain to which the service function belongs is selectively given by the service provider when the service is published. In the filtering phase based on the service category, if the service requester provides the service category constraint, the candidate service set is selected according to the category. Each service that includes a service category constraint is traversed, and when the service category is not included in the requested service category, the service is filtered out; when the candidate service set is filtered based on the number of service input and output, the premise hypothesis service output is introduced. When the number of parameters is less than the number of output parameters required by the query request message, it is considered that the service cannot provide the complete information required for the query request message, and the service and the query request message cannot be matched; when the number of service input parameters is greater than the input of the query request message When the number of numbers is considered, it is considered that the query request message cannot provide all the inputs required to satisfy the service, and the service also fails to match the query request message; based on the above description, the number of input parameters in the candidate service set is greater than the input parameter of the query request message. The services of the trees are filtered out; the services of the output parameters of the database are smaller than the number of output parameters of the query request message; in addition, due to any domain ontology, the description of the concept in the field is complete, different ontology descriptions Collar The domain is also different, so if the service does not reference the ontology referenced in the request, then the service cannot provide the same functional parameters as the functional parameters required by the query request message, and there is no possibility between the service and the query request message. Similarity relationship, so the services in the service candidate set that do not reference the referenced ontology can be filtered out by the above process.
上述步骤 206中， 服务功能性匹配方法具体为： 对于给定的服务输入功 能性矩阵集合
In the above step 206, the service functional matching method is specifically: inputting a functional matrix set for a given service
FRi2… FRik} , FRo={FRoi , FRo2… FRok}： Result= {iResult j , iResult2… iResultk}={Fi1"1*FRi1 ,
}(1 < m < k), 其中， iResult由 k个矩阵组成， k为服务 S和查询请求消息 R输入参数所有引用本 体的数目中的最大值； 上述 k个本体构成服务匹配阶段的本体集合； 对于任 意本体 ontj , 若有服务 S或查询请求消息 R的输出参数没用引用 οη , 则对 应的服务功能性矩阵集合或者请求功能性矩阵集合的第 j 个矩阵元素用全零 替代； iResult中第 m个元素为 Fim ^FRi^ 表示服务 S的输入功能性矩阵 Fim 与查询请求消息 R的输入功能性矩阵 FRi„^ 乘积。 FRi 2 ... FRi k } , FRo={FRoi , FRo 2 ... FRo k }: Result= {iResult j , iResult 2 ... iResult k }={Fi 1 " 1 *FRi 1 , }(1 < m < k), where iResult consists of k matrices, k is the maximum of the number of all referenced ontology of the service S and query request message R input parameters; the above k ontology constitutes the ontology set of the service matching phase For any ontology ontj, if the output parameter of service S or query request message R is not referenced οη, then the corresponding service functional matrix set or the jth matrix element of the request functional matrix set is replaced with all zeros; iResult The mth element is Fi m ^FRi^ which represents the product of the input functional matrix Fi m of the service S and the input functional matrix FRi „^ of the query request message R.
上述步骤 207中， 服务非功能性匹配方法具体为： 接受服务服务候选集 合中任意服务 S与查询请求消息 R, 计算并返回服务 S与查询请求消息 R的 非功能属性匹配相似度； 对于 查询请求消息 R 中定义的每一个服务质量 ( QoS )的属性 qo ,判断其在服务 S中是否也定义了相同的属性 qo ,若有， 则计算服务 S中的该属性的取值范围与查询请求消息 R中该属性取值范围重 叠部分占查询请求消息 R中该属性取值范围的比值， 该比值为服务 S与查询 请求消息 R在 qosj属性上的相似度； 最后通过计算所有 QoS属性上相似度的 平均值来计算服务 S与查询请求消息 R的非功能匹配相似度。 In the foregoing step 207, the service non-functional matching method is specifically: accepting any service S and the query request message R in the service service candidate set, and calculating and returning the non-functional attribute matching similarity between the service S and the query request message R; The attribute qo of each quality of service (QoS) defined in the message R is judged whether it also defines the same attribute qo in the service S, and if so, the value range and the query request message of the attribute in the service S are calculated. In R, the overlap of the attribute value range occupies the ratio of the value range of the attribute in the query request message R, and the ratio is the similarity between the service S and the query request message R on the qosj attribute; finally, the similarity of all QoS attributes is calculated. The average value is used to calculate the non-functional matching similarity between the service S and the query request message R.
上述步骤 208中，根据服务的文本特征向量 V和请求的文本特征向量 ν' , 通过公式 ts=v X v'计算服务和请求的文本特征相似度 ts; In the above step 208, according to the text feature vector V of the service and the requested text feature vector ν', the text feature similarity ts of the service and the request is calculated by the formula ts=v X v';
上述步骤 209中， 对功能相似度矩阵、 非功能相似度矩阵和文本相似度 矩阵进行融合， 按照最后的相似度对服务进行排序， 根据功能相似度 ms, 非
功能相似度 qs和文本相似度 ts查询请求消息 R和服务 的相似度 S(r, Ci) , 具体地， 如公式 S(r, )=a x ms+(l-a) X qs+ts所示， 其中， 参数 a(0<a<l)是一 个用户可调节的参数， 该参数 a反映了用户对功能和非功能性的偏好， 优选 地， 取参数 a=0.7。 In the above step 209, the functional similarity matrix, the non-functional similarity matrix, and the text similarity matrix are merged, and the services are sorted according to the final similarity, according to the functional similarity ms, The functional similarity qs and the text similarity ts query the similarity S(r, Ci) of the request message R and the service, specifically, as shown by the formula S(r, )=ax ms+(la) X qs+ts , where The parameter a (0 < a < l) is a user adjustable parameter that reflects the user's preference for functional and non-functionality, preferably taking the parameter a = 0.7.
本发明实施例提供的基于多特征匹配的服务发现方法， 根据请求输入功 能性矩阵集合和请求输出功能性矩阵集合计算与查询请求消息相关联的候选 服务集合中的每一个服务之间的相似度， 并根据相似度获取与查询请求消息 待查询的服务列表， 由于在获取服务时避免了对服务的语义进行推理以及对 服务的相似度进行计算， 从而提高了服务与请求查询消息的匹配效率， 进一 步提高了获取服务的时间效率。 The multi-feature matching based service discovery method provided by the embodiment of the present invention calculates the similarity between each service in the candidate service set associated with the query request message according to the request input functional matrix set and the request output functional matrix set. And obtaining the service list to be queried according to the similarity degree, and avoiding the reasoning of the semantics of the service and the similarity of the service when the service is acquired, thereby improving the matching efficiency between the service and the request query message, The time efficiency of obtaining services is further improved.
图 3 为本发明基于多特征匹配的服务发现***一个实施例的结构示意 图， 如图 3所示， 本发明实施例包括： 第一获耳 莫块 31、 计算模块 32、 第二 获取模块 33; FIG. 3 is a schematic structural diagram of an embodiment of a service discovery system based on multi-feature matching according to the present invention. As shown in FIG. 3, the embodiment of the present invention includes: a first acquisition module 31, a calculation module 32, and a second acquisition module 33;
其中，第一获取模块 31根据查询请求消息获取请求输入功能性矩阵集合 和请求输出功能性矩阵集合；计算模块 32根据所述请求输入功能性矩阵集合 和请求输出功能性矩阵集合计算与所述查询请求消息相关联的候选服务集合 中的每一个服务之间的相似度；第二获取模块 33根据所述相似度获取所述查 询请求消息待查询的服务列表。 The first obtaining module 31 acquires a request input functional matrix set and a request output functional matrix set according to the query request message; the calculating module 32 calculates and the query according to the request input functional matrix set and the request output functional matrix set. The similarity between each of the candidate service sets associated with the request message; the second obtaining module 33 acquires the service list to be queried by the query request message according to the similarity.
本发明实施例提供的基于多特征匹配的服务发现***， 通过计算模块 32 根据请求输入功能性矩阵集合和请求输出功能性矩阵集合计算与查询请求消 息相关联的候选服务集合中的每一个服务之间的相似度，第二获取模块 33根 据相似度获取与查询请求消息待查询的服务列表， 由于在获取服务时避免了 对服务的语义进行推理以及对服务的相似度进行计算， 从而提高了服务与请 求查询消息的匹配效率， 进一步提高了获取服务的时间效率。 The multi-feature matching based service discovery system provided by the embodiment of the present invention calculates, by the calculation module 32, each service in the candidate service set associated with the query request message according to the request input functional matrix set and the request output functional matrix set. The second acquisition module 33 obtains the service list to be queried according to the similarity degree according to the similarity degree, and improves the service by avoiding the reasoning of the semantics of the service and calculating the similarity of the service when acquiring the service. The efficiency of matching with the request query message further improves the time efficiency of obtaining the service.
图 4为本发明基于多特征匹配的服务发现***又一个实施例的结构示意 图， 如图 4所示， 本发明实施例包括： 第一获取模块 41、 计算模块 42、 第二
获取模块 43; 4 is a schematic structural diagram of still another embodiment of a service discovery system based on multi-feature matching according to the present invention. As shown in FIG. 4, the embodiment of the present invention includes: a first obtaining module 41, a calculating module 42, and a second Obtaining module 43;
其中，第一获取模块 41根据查询请求消息获取请求输入功能性矩阵集合 和请求输出功能性矩阵集合；计算模块 42根据所述请求输入功能性矩阵集合 和请求输出功能性矩阵集合计算与所述查询请求消息相关联的候选服务集合 中的每一个服务之间的相似度；第二获取模块 43根据所述相似度获取所述查 询请求消息待查询的服务列表。 The first obtaining module 41 acquires a request input functional matrix set and a request output functional matrix set according to the query request message; the calculating module 42 calculates and the query according to the request input functional matrix set and the request output functional matrix set. The similarity between each service in the candidate service set associated with the request message; the second obtaining module 43 acquires the service list to be queried by the query request message according to the similarity.
进一步地， 第一获取模块 41还可以包括： 抽取单元 411、 第一获取单元 412、 第二获取单元 413; 其中， 抽取单元 411对所述查询请求消息相对应的 发布文档进行语义信息抽取， 得到所述查询请求消息所有本体的数目、 所述 查询请求消息的输入参数与所述查询请求消息引用的多个本体中所有输入参 数语义概念的关系； 第一获取单元 412根据所述所有本体的数目、 所述查询 请求消息的输入参数、 所述查询请求消息的输出参数与所述查询请求消息引 用的多个本体中所有输入参数语义概念的关系获取请求输入功能性矩阵集 合； 第二获取单元 413根据所述所有本体的数目、 所述查询请求消息的输出 参数与所述查询请求消息引用的多个本体中所有输出参数语义概念的关系获 取请求输出功能性矩阵集合。 Further, the first obtaining module 41 may further include: an extracting unit 411, a first obtaining unit 412, and a second obtaining unit 413; wherein, the extracting unit 411 performs semantic information extraction on the published document corresponding to the query request message, to obtain The relationship between the number of all the ontology of the query request message, the input parameter of the query request message, and the semantic concept of all the input parameters in the plurality of ontology referenced by the query request message; the first obtaining unit 412 is configured according to the number of all the ontology The input parameter of the query request message, the output parameter of the query request message, and the relationship between all the input parameter semantic concepts in the plurality of ontology referenced by the query request message acquire the request input functional matrix set; the second obtaining unit 413 And obtaining a request output functional matrix set according to the relationship between the number of all the ontology, the output parameter of the query request message, and the semantic concepts of all output parameters in the plurality of ontology referenced by the query request message.
进一步地， 计算模块 42还可以包括： 第一过滤单元 421、 第二过滤单元 422、 第一计算单元 423; 其中， 第一过滤单元 421根据所述查询请求消息对 保存在数据库中的多个服务进行过滤； 第二过滤单元 422根据所述多个服务 的类型信息、 所述多个服务的输入输出数量以及服务 ^ I用到的本体信息对服 务进行过滤， 得到与所述查询请求消息相关联的候选服务集合； 第一计算单 元 423 计算所述候选服务集合中的每一个服务与所述查询请求消息的相似 度。 Further, the calculation module 42 may further include: a first filtering unit 421, a second filtering unit 422, and a first calculating unit 423; wherein, the first filtering unit 421 searches for multiple services stored in the database according to the query request message. Performing filtering; the second filtering unit 422 filters the service according to the type information of the multiple services, the number of input and output of the multiple services, and the ontology information used by the service, to obtain the association with the query request message. a candidate service set; the first computing unit 423 calculates a similarity of each of the candidate service sets to the query request message.
进一步地， 第一计算单元 423还可以包括： 第一计算子单元、 第二计算 子单元、 第一获取子单元、 第二获取子单元； 其中， 第一计算子单元根据服 务功能性匹配方法计算所述候选服务集合中的每一个服务和所述查询请求消
息的功能相似度矩阵； 第二计算子单元根据服务非功能性匹配方法计算所述 候选服务集合中的每一个服务和所述查询请求消息的非功能相似度矩阵； 第 一获取子单元根据所述候选服务集合中的每一个服务的文本特征向量和所述 查询请求消息的文本特征向量获取文本相似度矩阵； 第二获取子单元根据所 述功能相似度矩阵、 非功能相似度矩阵和文本相似度矩阵获取所述查询请求 消息与所述候选服务中的每一个服务的相似度。 Further, the first calculating unit 423 may further include: a first calculating subunit, a second calculating subunit, a first acquiring subunit, and a second acquiring subunit; wherein the first calculating subunit is calculated according to a service functional matching method Each of the candidate service sets and the query request cancellation a functional similarity matrix; the second computing subunit calculates a non-functional similarity matrix of each of the candidate service sets and the query request message according to a service non-functional matching method; a text feature vector of each service in the candidate service set and a text feature vector of the query request message to obtain a text similarity matrix; the second obtaining subunit is based on the functional similarity matrix, the non-functional similarity matrix, and the text similarity The degree matrix obtains the similarity between the query request message and each of the candidate services.
进一步地， 第二获取子单元还可以包括： 乘法子单元和计算子单元； 其 中， 乘法子单元， 用于根据功能相似度矩阵和非功能相似度矩阵进行矩阵乘 积， 生成输入匹配结果矩阵； 计算子单元， 用于计算所述匹配结果矩阵中所 有非零元素的数目，该所述所有非零元素的数目与所有输入参数和输出参数 的个数的比值作为所述服务与所述查询请求消息匹配的相似度。 Further, the second obtaining subunit may further include: a multiplication subunit and a calculation subunit; wherein, the multiplication subunit is configured to perform matrix matching on the functional similarity matrix and the non-functional similarity matrix to generate an input matching result matrix; a subunit, configured to calculate a number of all non-zero elements in the matching result matrix, the ratio of the number of all non-zero elements to the number of all input parameters and output parameters as the service and the query request message The similarity of the match.
本发明实施例提供的基于多特征匹配的服务发现方法， 通过计算模块 42 根据请求输入功能性矩阵集合和请求输出功能性矩阵集合计算与查询请求消 息相关联的候选服务集合中的每一个服务之间的相似度，第二获取模块 43根 据相似度获取与查询请求消息待查询的服务列表， 由于在获取服务时避免了 对服务的语义进行推理以及对服务的相似度进行计算， 从而提高了服务与请 求查询消息的匹配效率， 进一步提高了获取服务的时间效率。 The multi-feature matching based service discovery method provided by the embodiment of the present invention calculates, by the calculation module 42, each service in the candidate service set associated with the query request message according to the request input functional matrix set and the request output functional matrix set. The second obtaining module 43 obtains the service list to be queried according to the similarity degree according to the similarity degree, and improves the service because the semantics of the service are avoided and the similarity of the service is calculated when the service is acquired. The efficiency of matching with the request query message further improves the time efficiency of obtaining the service.
所属领域的技术人员可以清楚地了解到， 为描述的方便和简洁， 上述描 述的***、 设备、 模块和单元的具体工作过程， 可以参考前述方法实施例中 的对应过程， 在此不再赘述。 A person skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working process of the system, the device, the module and the unit described above can refer to the corresponding process in the foregoing method embodiment, and details are not described herein again.
本领域普通技术人员可以理解： 实现上述实施例的全部或部分步骤可以 通过程序指令相关的硬件来完成， 前述的程序可以存储于一计算机可读取存 储介质中， 该程序在执行时， 执行包括上述方法实施例的步骤； 而前述的存 储介质包括： ROM、 RAM, 磁碟或者光盘等各种可以存储程序代码的介质。 It will be understood by those skilled in the art that all or part of the steps of implementing the foregoing embodiments may be performed by hardware related to program instructions. The foregoing program may be stored in a computer readable storage medium, and when executed, the program includes The foregoing steps of the method embodiment; and the foregoing storage medium includes: a medium that can store program codes, such as a ROM, a RAM, a magnetic disk, or an optical disk.
最后应说明的是： 以上实施例仅用以说明本发明的技术方案， 而非对其 限制； 尽管参照前述实施例对本发明进行了详细的说明， 本领域的普通技术
人员应当理解： 其依然可以对前述各实施例所记载的技术方案进行修改， 或 者对其中部分技术特征进行等同替换； 而这些修改或者替换， 并不使相应技 术方案的本质脱离本发明各实施例技术方案的精神和范围。
Finally, it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, and are not intended to be limiting thereof; although the present invention has been described in detail with reference to the foregoing embodiments, It should be understood that: the technical solutions described in the foregoing embodiments may be modified, or some of the technical features may be equivalently replaced; and the modifications or substitutions do not deviate from the essence of the corresponding technical solutions. The spirit and scope of the technical programme.
Claims
1、 一种基于多特征匹配的服务发现方法， 其特征在于， 包括： 根据查询请求消息获取请求输入功能性矩阵集合和请求输出功能性矩阵 根据所述请求输入功能性矩阵集合和请求输出功能性矩阵集合计算与所 述查询请求消息相关联的候选服务集合中的每一个服务之间的相似度； A service discovery method based on multi-feature matching, comprising: obtaining a functional matrix set and a request output functional matrix according to a query request message acquisition request function matrix set and request output function according to the request The matrix set calculates a similarity between each of the candidate service sets associated with the query request message;
根据所述相似度获取所述查询请求消息待查询的服务列表。 Obtaining, according to the similarity, a service list to be queried by the query request message.
2、 根据权利要求 1所述的方法， 其特征在于， 所述根据查询请求消息获 取请求输入功能性矩阵集合和请求输出功能性矩阵集合包括： The method according to claim 1, wherein the obtaining the request input functional matrix set and the request output functional matrix set according to the query request message comprises:
对查询请求消息相对应的发布文档进行语义信息抽取， 得到所述查询请 求消息所有本体的数目、 所述查询请求消息的输入参数与所述查询请求消息 引用的多个本体中所有输入参数语义概念的关系； Semantic information extraction is performed on the published document corresponding to the query request message, and the number of all the ontology of the query request message, the input parameter of the query request message, and the semantic concept of all input parameters in the plurality of ontology referenced by the query request message are obtained. Relationship;
根据所述所有本体的数目、 所述查询请求消息的输入参数、 所述查询请 求消息的输出参数与所述查询请求消息引用的多个本体中所有输入参数语义 概念的关系获取请求输入功能性矩阵集合； Obtaining a request input functional matrix according to the relationship between the number of all the ontology, the input parameter of the query request message, the output parameter of the query request message, and the semantic concept of all input parameters in the plurality of ontology referenced by the query request message Collection
根据所述所有本体的数目、 所述查询请求消息的输出参数与所述查询请 求消息引用的多个本体中所有输出参数语义概念的关系获取请求输出功能性 矩阵集合。 And obtaining a request output functional matrix set according to the relationship between the number of all the ontology, the output parameter of the query request message, and the semantic concepts of all output parameters in the plurality of ontology referenced by the query request message.
3、 根据权利要求 1或 2所述的方法， 其特征在于， 所述根据所述请求输 入功能性矩阵集合和请求输出功能性矩阵集合计算与所述查询请求消息相关 联的候选服务集合中的每一个服务之间的相似度包括： The method according to claim 1 or 2, wherein the calculating, according to the request input functional matrix set and the request output functional matrix set, the candidate service set associated with the query request message The similarities between each service include:
根据所述查询请求消息对保存在数据库中的多个服务进行过滤； 根据所述多个服务的类型信息、 所述多个服务的输入输出数量以及服务 引用到的本体信息对服务进行过滤， 得到与所述查询请求消息相关联的候选 服务集合； And filtering the plurality of services stored in the database according to the query request message; filtering the service according to the type information of the multiple services, the number of input and output of the multiple services, and the ontology information referenced by the service, to obtain a set of candidate services associated with the query request message;
计算所述候选服务集合中的每一个服务与所述查询请求消息的相似度。 Calculating a similarity of each of the candidate service sets to the query request message.
4、 根据权利要求 3所述的方法， 其特征在于， 所述计算所述候选服务集 合中的每一个服务与所述查询请求消息的相似度包括： The method according to claim 3, wherein the calculating the similarity between each of the candidate service sets and the query request message comprises:
根据服务功能性匹配方法计算所述候选服务集合中的每一个服务和所述 查询请求消息的功能相似度矩阵； Calculating a functional similarity matrix of each of the candidate service sets and the query request message according to a service functional matching method;
根据服务非功能性匹配方法计算所述候选服务集合中的每一个服务和所 述查询请求消息的非功能相似度矩阵； Calculating a non-functional similarity matrix of each of the candidate service sets and the query request message according to a service non-functional matching method;
根据所述候选服务集合中的每一个服务的文本特征向量和所述查询请求 消息的文本特征向量获取文本相似度矩阵； Obtaining a text similarity matrix according to a text feature vector of each of the candidate service sets and a text feature vector of the query request message;
根据所述功能相似度矩阵、 非功能相似度矩阵和文本相似度矩阵获取所 述查询请求消息与所述候选服务中的每一个服务的相似度。 Obtaining a similarity between the query request message and each of the candidate services according to the functional similarity matrix, the non-functional similarity matrix, and the text similarity matrix.
5、 根据权利要求 4所述的方法， 其特征在于， 所述根据所述功能相似度 矩阵、 非功能相似度矩阵和文本相似度矩阵获取所述查询请求消息与所述候 选服务中的每一个服务的相似度包括： The method according to claim 4, wherein the acquiring the query request message and each of the candidate services according to the functional similarity matrix, the non-functional similarity matrix, and the text similarity matrix Service similarities include:
根据功能相似度矩阵和非功能相似度矩阵进行矩阵乘积， 生成输入匹配 结果矩阵； Generating a matrix of input matching results according to a matrix product of a functional similarity matrix and a non-functional similarity matrix;
计算所述匹配结果矩阵中所有非零元素的数目，该所述所有非零元素的 数目与所有输入参数和输出参数的个数的比值作为所述服务与所述查询请求 消息匹配的相似度。 A number of all non-zero elements in the matching result matrix is calculated, the ratio of the number of all non-zero elements to the number of all input parameters and output parameters as the similarity of the service to the query request message.
6、 一种能够实现权利要求 1 ~ 5任一所述方法的基于多特征匹配的服务 发现***， 其特征在于， 包括： 6. A multi-feature matching based service discovery system capable of implementing the method of any one of claims 1 to 5, comprising:
第一获取模块， 用于根据查询请求消息获取请求输入功能性矩阵集合和 请求输出功能性矩阵集合； a first acquiring module, configured to acquire a request input functional matrix set and a request output functional matrix set according to the query request message;
计算模块， 用于根据所述请求输入功能性矩阵集合和请求输出功能性矩 阵集合计算与所述查询请求消息相关联的候选服务集合中的每一个服务之间 的相似度； a calculation module, configured to calculate, according to the request input functional matrix set and the request output functional matrix set, a similarity between each of the candidate service sets associated with the query request message;
第二获取模块， 用于根据所述相似度获取所述查询请求消息待查询的服 务列表。 a second acquiring module, configured to acquire, according to the similarity, the query request message to be queried List of services.
7、 根据权利要求 6所述的***， 其特征在于， 所述第一获取模块包括： 抽取单元， 用于对所述查询请求消息相对应的发布文档进行语义信息抽 取， 得到所述查询请求消息所有本体的数目、 所述查询请求消息的输入参数 与所述查询请求消息引用的多个本体中所有输入参数语义概念的关系； The system according to claim 6, wherein the first obtaining module comprises: an extracting unit, configured to perform semantic information extraction on a publishing document corresponding to the query request message, to obtain the query request message. The relationship between the number of all the ontology, the input parameter of the query request message, and the semantic concept of all input parameters in the plurality of ontology referenced by the query request message;
第一获取单元， 用于根据所述所有本体的数目、 所述查询请求消息的输 入参数、 所述查询请求消息的输出参数与所述查询请求消息引用的多个本体 中所有输入参数语义概念的关系获取请求输入功能性矩阵集合； a first acquiring unit, configured to: according to the number of all the ontology, an input parameter of the query request message, an output parameter of the query request message, and a semantic concept of all input parameters in the plurality of ontology referenced by the query request message Relationship acquisition request input functional matrix set;
第二获取单元， 用于根据所述所有本体的数目、 所述查询请求消息的输 出参数与所述查询请求消息引用的多个本体中所有输出参数语义概念的关系 获取请求输出功能性矩阵集合。 And a second obtaining unit, configured to obtain a request output functional matrix set according to the relationship between the number of all the ontology, the output parameter of the query request message, and the semantic concepts of all output parameters in the plurality of ontology referenced by the query request message.
8、 根据权利要求 6或 7所述的***， 其特征在于， 所述计算模块包括： 第一过滤单元， 用于根据所述查询请求消息对保存在数据库中的多个服 务进行过滤； The system according to claim 6 or 7, wherein the calculation module comprises: a first filtering unit, configured to filter, according to the query request message, a plurality of services stored in a database;
第二过滤单元， 用于根据所述多个服务的类型信息、 所述多个服务的输 入输出数量以及服务引用到的本体信息对服务进行过滤， 得到与所述查询请 求消息相关联的候选服务集合； a second filtering unit, configured to filter the service according to the type information of the multiple services, the number of input and output of the multiple services, and the ontology information referenced by the service, to obtain a candidate service associated with the query request message Collection
第一计算单元， 用于计算所述候选服务集合中的每一个服务与所述查询 请求消息的相似度。 And a first calculating unit, configured to calculate a similarity between each of the candidate service sets and the query request message.
9、 根据权利要求 8所述的***， 其特征在于， 所述第一计算单元包括： 第一计算子单元， 用于根据服务功能性匹配方法计算所述候选服务集合 中的每一个服务和所述查询请求消息的功能相似度矩阵； The system according to claim 8, wherein the first calculating unit comprises: a first calculating subunit, configured to calculate each service and location in the candidate service set according to a service functional matching method a functional similarity matrix of the query request message;
第二计算子单元， 用于根据服务非功能性匹配方法计算所述候选服务集 合中的每一个服务和所述查询请求消息的非功能相似度矩阵； a second calculating subunit, configured to calculate, according to a service non-functional matching method, each of the candidate service sets and a non-functional similarity matrix of the query request message;
第一获取子单元， 用于根据所述候选服务集合中的每一个服务的文本特 征向量和所述查询请求消息的文本特征向量获取文本相似度矩阵； 第二获取子单元， 用于根据所述功能相似度矩阵、 非功能相似度矩阵和 文本相似度矩阵获取所述查询请求消息与所述候选服务中的每一个服务的相 似度。 a first obtaining subunit, configured to obtain a text similarity matrix according to a text feature vector of each service in the candidate service set and a text feature vector of the query request message; And a second acquiring subunit, configured to acquire, according to the functional similarity matrix, the non-functional similarity matrix, and the text similarity matrix, a similarity between the query request message and each of the candidate services.
10、 根据权利要求 9所述的***， 其特征在于， 所述第二获取子单元包 括： 10. The system according to claim 9, wherein the second obtaining subunit comprises:
乘法子单元， 用于根据功能相似度矩阵和非功能相似度矩阵进行矩阵乘 积， 生成输入匹配结果矩阵； a multiplication subunit, configured to perform matrix multiplication according to a functional similarity matrix and a non-functional similarity matrix to generate an input matching result matrix;
计算子单元，用于计算所述匹配结果矩阵中所有非零元素的数目，该所述 所有非零元素的数目与所有输入参数和输出参数的个数的比值作为所述服务 与所述查询请求消息匹配的相似度。 a calculation subunit, configured to calculate a number of all non-zero elements in the matching result matrix, the ratio of the number of all non-zero elements to the number of all input parameters and output parameters as the service and the query request The similarity of message matching.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110047821.7A CN102651755B (en) | 2011-02-28 | 2011-02-28 | Service discovering method and system based on multi-feature matching |
| CN201110047821.7 | 2011-02-28 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2012116512A1 true WO2012116512A1 (en) | 2012-09-07 |
Family
ID=46693649
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2011/072647 WO2012116512A1 (en) | 2011-02-28 | 2011-04-12 | Service discovery method and system based on multiple features matching |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN102651755B (en) |
| WO (1) | WO2012116512A1 (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113434656A (en) * | 2021-07-21 | 2021-09-24 | 广州华多网络科技有限公司 | E-commerce customer service matching method and corresponding device, equipment and medium thereof |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106529801B (en) * | 2016-11-09 | 2019-11-01 | 北京航空航天大学 | Services Composition Candidate Set construction method and system |
| CN109948062A (en) * | 2019-03-29 | 2019-06-28 | 广州市百果园网络科技有限公司 | A kind of target matching method, device, server, system and storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101674290A (en) * | 2008-11-26 | 2010-03-17 | 天津大学 | Semantics-based automatic service combination system for web service relation network |
| US7877726B2 (en) * | 2004-07-13 | 2011-01-25 | The Mitre Corporation | Semantic system for integrating software components |
-
2011
- 2011-02-28 CN CN201110047821.7A patent/CN102651755B/en not_active Expired - Fee Related
- 2011-04-12 WO PCT/CN2011/072647 patent/WO2012116512A1/en active Application Filing
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7877726B2 (en) * | 2004-07-13 | 2011-01-25 | The Mitre Corporation | Semantic system for integrating software components |
| CN101674290A (en) * | 2008-11-26 | 2010-03-17 | 天津大学 | Semantics-based automatic service combination system for web service relation network |
Non-Patent Citations (3)
| Title |
|---|
| SONG LING ET AL.: "Semantic Similarity Computation of Concepts and Documents.", COMPUTER ENGINEERING AND APPLICATIONS., vol. 44, no. 35, 11 December 2008 (2008-12-11) * |
| WU JIAN ET AL.: "Web Service Discovery Based on Ontology and Similarity of Words.", CHINESE JOURNAL OF COMPUTER., vol. 28, no. 4, 12 April 2005 (2005-04-12) * |
| ZHANG XIAN ET AL.: "An Approach and Implementation to Semantic Web Service Matchmaking.", COMPUTER SCIENCE., vol. 34, no. 5, 25 May 2007 (2007-05-25) * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113434656A (en) * | 2021-07-21 | 2021-09-24 | 广州华多网络科技有限公司 | E-commerce customer service matching method and corresponding device, equipment and medium thereof |
| CN113434656B (en) * | 2021-07-21 | 2023-04-25 | 广州华多网络科技有限公司 | E-commerce customer service matching method and corresponding device, equipment and medium thereof |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102651755B (en) | 2015-02-18 |
| CN102651755A (en) | 2012-08-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2020007224A1 (en) | Knowledge graph construction and smart response method and apparatus, device, and storage medium | |
| US9372938B2 (en) | Augmenting queries when searching a semantic database | |
| TWI496015B (en) | Text matching method and device | |
| KR100815563B1 (en) | System and method for knowledge extension and inference service based on DBMS | |
| Gilson et al. | From web data to visualization via ontology mapping | |
| CN105760397B (en) | Internet of things ontology model processing method and device | |
| TW201248534A (en) | Method and system of recommending items | |
| US9753960B1 (en) | System, method, and computer program for dynamically generating a visual representation of a subset of a graph for display, based on search criteria | |
| CN106469169A (en) | Information processing method and device | |
| WO2017166944A1 (en) | Method and device for providing service access | |
| CN102289473A (en) | Device and method for paging query of a plurality of servers | |
| WO2011011063A2 (en) | Method and system for document indexing and data querying | |
| US20190387056A1 (en) | Irc-infoid data standardization for use in a plurality of mobile applications | |
| Zhang et al. | Dual personalization on federated recommendation | |
| Wang et al. | PCSG: pattern-coverage snippet generation for RDF datasets | |
| WO2012116512A1 (en) | Service discovery method and system based on multiple features matching | |
| CN109903191A (en) | Travelling recommended method, device, storage medium and terminal based on machine learning | |
| CN112330400B (en) | Product recommendation method and device, computer equipment and storage medium | |
| KR20210063878A (en) | A method and an apparatus for providing chatbot services of analyzing marketing information | |
| CN108959537A (en) | Web service interface acquisition methods and device | |
| Rupasingha et al. | Calculating Web service similarity using ontology learning with machine learning | |
| CN106055699B (en) | A kind of method and device of feature clustering | |
| CN111242146B (en) | POI information classification based on convolutional neural network | |
| WO2014117296A1 (en) | Generating a hint for a query | |
| Li et al. | Building an Open Cloud Virtual Dataspace Model for Materials Scientific Data. |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 11860047 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 11860047 Country of ref document: EP Kind code of ref document: A1 |