【摘要】：Web服務(wù)技術(shù)不僅是解決軟件重用問題的重要解決方案,而且還是復(fù)雜異構(gòu)軟件系統(tǒng)低耦合設(shè)計(jì)的理論基礎(chǔ)。Web服務(wù)發(fā)現(xiàn)作為Web服務(wù)技術(shù)的一個(gè)核心過程,是實(shí)現(xiàn)Web服務(wù)組合、Web服務(wù)調(diào)用等其他過程的基礎(chǔ)。開放網(wǎng)絡(luò)中不斷增加的Web服務(wù)為工業(yè)生產(chǎn)帶來了便利,但也對(duì)服務(wù)發(fā)現(xiàn)系統(tǒng)在查全率、查準(zhǔn)率和響應(yīng)時(shí)間三個(gè)方面提出了新的挑戰(zhàn)。不僅如此,由于服務(wù)發(fā)現(xiàn)系統(tǒng)中存在著大量功能相同但質(zhì)量參差不齊的Web服務(wù),如何對(duì)服務(wù)進(jìn)行非功能屬性(QoS, Quality of Service)篩選將最優(yōu)質(zhì)的服務(wù)提供給用戶也成為了服務(wù)發(fā)現(xiàn)系統(tǒng)的另一個(gè)問題。 以上問題在目前Web服務(wù)發(fā)現(xiàn)系統(tǒng)中表現(xiàn)為單點(diǎn)故障、低效的資源定位方法、低效的網(wǎng)絡(luò)資源利用方法、低精度的服務(wù)匹配方法、缺少對(duì)QoS信息的篩選與維護(hù)等,為此本文提出了P2P網(wǎng)絡(luò)環(huán)境下支持QoS篩選的語(yǔ)義Web服務(wù)發(fā)現(xiàn)框架。本文對(duì)Chord網(wǎng)絡(luò)模型與JXTA分組方法進(jìn)行了研究,將系統(tǒng)中的網(wǎng)絡(luò)節(jié)點(diǎn)組成邏輯上的雙層架構(gòu),并通過兩階段匹配方法在進(jìn)行服務(wù)匹配步驟前過濾掉大部分不符合要求的服務(wù)節(jié)點(diǎn),再輔以Chord模型的快速路由將服務(wù)請(qǐng)求轉(zhuǎn)發(fā)至目標(biāo)節(jié)點(diǎn),以此實(shí)現(xiàn)服務(wù)資源的快速定位。本文對(duì)服務(wù)匹配方法進(jìn)行了探討,提出了基于WordNet本體結(jié)構(gòu)的概念語(yǔ)義相似度匹配算法,該算法實(shí)現(xiàn)了對(duì)基于自然語(yǔ)言的服務(wù)請(qǐng)求以及服務(wù)描述的語(yǔ)義匹配支持,降低了對(duì)用戶使用服務(wù)發(fā)現(xiàn)系統(tǒng)的要求,提高了服務(wù)匹配過程的性能。本文研究了現(xiàn)有服務(wù)發(fā)現(xiàn)系統(tǒng)對(duì)QoS信息的設(shè)定和使用,提出了完整的QoS信息采集與維護(hù)的流程,實(shí)現(xiàn)了對(duì)Web服務(wù)QoS匹配的支持。本文分析了現(xiàn)有服務(wù)發(fā)現(xiàn)的工作流程,設(shè)計(jì)了歷史管理模塊和資源管理模塊來進(jìn)行優(yōu)化。通過對(duì)節(jié)點(diǎn)的歷史數(shù)據(jù)的挖掘,可以實(shí)現(xiàn)基于用戶偏好的服務(wù)發(fā)現(xiàn),當(dāng)用戶的選擇具有歷史傾向性時(shí)能極大的縮短服務(wù)發(fā)現(xiàn)的響應(yīng)時(shí)間；而資源管理模塊收集的節(jié)點(diǎn)剩余資源可以用來計(jì)算每個(gè)節(jié)點(diǎn)的綜合處理能力值,為超級(jí)節(jié)點(diǎn)和影子節(jié)點(diǎn)的選擇和QoS驗(yàn)證的調(diào)度提供了數(shù)據(jù)支持,這樣就不僅實(shí)現(xiàn)了對(duì)網(wǎng)絡(luò)資源的充分利用,還提高了服務(wù)發(fā)現(xiàn)系統(tǒng)的穩(wěn)定性。 本文設(shè)計(jì)了兩個(gè)實(shí)驗(yàn),分別對(duì)概念語(yǔ)義相似度匹配算法和智能Web服務(wù)發(fā)現(xiàn)框架(SWSDI, Smart Web Service Discovery Infrastructure)服務(wù)發(fā)現(xiàn)系統(tǒng)進(jìn)行了性能驗(yàn)證。實(shí)驗(yàn)一的結(jié)果表明本文提出的改進(jìn)算法具有良好的穩(wěn)定性,并且在算法準(zhǔn)確性上要優(yōu)于其他算法；實(shí)驗(yàn)二則驗(yàn)證了SWSDI服務(wù)發(fā)現(xiàn)系統(tǒng)無論在查全率還是在查準(zhǔn)率上都能滿足用戶的基本需求,并驗(yàn)證了概念語(yǔ)義相似度匹配算法在SWSDI系統(tǒng)中的性能。
[Abstract]:Web services technology is not only an important solution to the problem of software reuse, but also the theoretical basis of low coupling design of complex heterogeneous software systems. Web service discovery is a core process of Web services technology. Is the implementation of Web services composition, Web service invocation and other procedures. The increasing number of Web services in open networks brings convenience to industrial production, but it also poses new challenges to service discovery systems in three aspects: recall, recall and response time. Not only that, because there are a large number of Web services with the same functions but uneven quality in the service discovery system, How to provide the best service to users by non-functional attribute (QoS, Quality of Service) filter has become another problem in service discovery system. The above problems in the current Web service discovery system are as follows: single point fault, inefficient resource location method, inefficient network resource utilization method, low precision service matching method, lack of screening and maintenance of QoS information, etc. In this paper, a semantic Web service discovery framework supporting QoS filtering in P2P network environment is proposed. In this paper, the Chord network model and the JXTA packet method are studied. The network nodes in the system are formed into a logical two-layer architecture, and most of the service nodes that do not meet the requirements are filtered out by the two-stage matching method before the service matching step. Then the service request is forwarded to the target node by the fast routing of Chord model to realize the fast location of service resource. In this paper, the method of service matching is discussed, and a concept semantic similarity matching algorithm based on WordNet ontology structure is proposed. The algorithm realizes semantic matching support for service request and service description based on natural language. The requirement of service discovery system is reduced, and the performance of service matching process is improved. This paper studies the setup and use of QoS information in existing service discovery systems, proposes a complete flow of QoS information collection and maintenance, and realizes the support of QoS matching for Web services. This paper analyzes the workflow of service discovery, and designs a history management module and a resource management module to optimize. By mining the historical data of nodes, the service discovery based on user preference can be realized, and the response time of service discovery can be greatly shortened when the user's choice has historical tendency. The residual resources collected by resource management module can be used to calculate the comprehensive processing capacity of each node, which provides data support for the selection of super nodes and shadow nodes and the scheduling of QoS verification. This not only realizes the full use of network resources, but also improves the stability of service discovery system. In this paper, two experiments are designed to verify the performance of the conceptual semantic similarity matching algorithm and the intelligent Web services discovery framework (SWSDI, Smart Web Service Discovery Infrastructure) service discovery system. The results of experiment 1 show that the improved algorithm has good stability and is superior to other algorithms in the accuracy of the algorithm. The second experiment verifies that the SWSDI service discovery system can meet the basic needs of users in terms of recall rate and recall rate, and verifies the performance of conceptual semantic similarity matching algorithm in SWSDI system.
【學(xué)位授予單位】：北京交通大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類號(hào)】：TP393.09

【參考文獻(xiàn)】

相關(guān)期刊論文 前6條

1 鄧志鴻,唐世渭,張銘,楊冬青,陳捷;Ontology研究綜述[J];北京大學(xué)學(xué)報(bào)(自然科學(xué)版);2002年05期

2 歐偉杰;曾承;韓棟;彭智勇;劉洋;馬景燕;劉波;;基于語(yǔ)義相似性的Web服務(wù)匹配算法[J];計(jì)算機(jī)科學(xué);2012年01期

3 張付志;周立娜;趙偉偉;許立永;;基于QoS的Web服務(wù)發(fā)現(xiàn)方法[J];計(jì)算機(jī)工程;2010年17期

4 李小林;張力娜;張順利;;基于QoS監(jiān)控的Web服務(wù)選擇方法[J];計(jì)算機(jī)應(yīng)用與軟件;2012年03期

5 鄭曉霞;王建仁;;基于QoS的Web服務(wù)發(fā)現(xiàn)模型研究[J];情報(bào)科學(xué);2007年02期

6 劉志忠;王懷民;周斌;;一種雙層P2P結(jié)構(gòu)的語(yǔ)義服務(wù)發(fā)現(xiàn)模型[J];軟件學(xué)報(bào);2007年08期

，

本文編號(hào)：2266471