【摘要】：隨著數(shù)據(jù)挖掘知識的普及,數(shù)據(jù)挖掘已經(jīng)在社會各行各業(yè)的應(yīng)用迅速擴展開來。通俗來講,數(shù)據(jù)挖掘就是從大量的數(shù)據(jù)中,提取人們感興趣的,即正確的、未知的、有潛在價值的并最終能被用戶接收和利用的知識。數(shù)據(jù)挖掘已經(jīng)廣泛地應(yīng)用于各個領(lǐng)域,越來越多的學(xué)者開始關(guān)注于數(shù)據(jù)挖掘領(lǐng)域。隨著數(shù)據(jù)挖掘相關(guān)知識在越來越多的領(lǐng)域開始普及,其相關(guān)的實用工具也開始受到人們的重視。數(shù)據(jù)挖掘的研究重點也慢慢向數(shù)據(jù)挖掘系統(tǒng)工具方向轉(zhuǎn)移,越來越注重多種多樣的發(fā)現(xiàn)策略和技術(shù)的集成。數(shù)據(jù)挖掘工具的用戶主要集中在一些傳統(tǒng)的行業(yè)中,比如銀行、制造業(yè)、銷售等面臨轉(zhuǎn)型的行業(yè),數(shù)據(jù)挖掘工具也在不斷的產(chǎn)生和完善。基于上述的場景和需求,本文研究并設(shè)計了一套基于流程的數(shù)據(jù)挖掘建模工具。該工具提供了一種基于web服務(wù)的流程建模方式,旨在為流程開發(fā)者提供一個統(tǒng)一的建模平臺,減少開發(fā)者的工作量,滿足不同開發(fā)者的動態(tài)需求。基于流程的數(shù)據(jù)挖掘建模方式借助于web服務(wù),通過將第三方服務(wù)進行組裝形成新的流程來滿足開發(fā)者的任務(wù)需求。這種開發(fā)方式的優(yōu)點在于開發(fā)者不需要過多地了解數(shù)據(jù)挖掘的相關(guān)知識,完全屏蔽了底層數(shù)據(jù)挖掘服務(wù)實現(xiàn)的細節(jié),只需要調(diào)用相關(guān)的基于數(shù)據(jù)挖掘的web服務(wù),調(diào)用服務(wù)對外提供的接口,對這些服務(wù)進行組裝,就能實現(xiàn)數(shù)據(jù)挖掘流程的搭建。該工具可顯著提升建模效率,對IT人員也可以大大減輕原有的IT支持代價,提升調(diào)度和數(shù)據(jù)運作效率。開發(fā)者可以自定義建模過程中每個步驟的細節(jié),并能夠?qū)崟r跟蹤流程的執(zhí)行情況。為了進一步減少開發(fā)人員的工作量,本系統(tǒng)為開發(fā)者提供了用于流程搭建的服務(wù)庫。將數(shù)據(jù)挖掘流程中常用到的數(shù)據(jù)預(yù)處理、挖掘算法、算法評估等任務(wù)進行封裝,以服務(wù)模板的形式提供給開發(fā)人員調(diào)用。借助于系統(tǒng)提供的服務(wù)庫,開發(fā)人員可以完成從數(shù)據(jù)輸入、數(shù)據(jù)拼接、特征抽取、模型訓(xùn)練、模型評估、模型應(yīng)用完整的建模過程,顯著提升流程建模效率。開發(fā)人員在搭建數(shù)據(jù)挖掘流程時,可以選擇導(dǎo)入第三方服務(wù),也可以調(diào)用系統(tǒng)提供服務(wù),極大地增加了流程開發(fā)的靈活性。本文首先闡述了數(shù)據(jù)挖掘建模系統(tǒng)的背景和意義,并研究系統(tǒng)所設(shè)計的相關(guān)技術(shù)和領(lǐng)域。接著從本文的整體目標出發(fā),分析了數(shù)據(jù)挖掘建模系統(tǒng)的功能性需求,根據(jù)需求對系統(tǒng)的總體架構(gòu)和模塊間的交互接口進行了設(shè)計,針對核心模塊的實現(xiàn)進行詳細介紹。最后對系統(tǒng)進行了測試,以驗證系統(tǒng)的合理性。
[Abstract]:With the popularization of data mining knowledge, data mining has been widely used in various industries. Generally speaking, data mining is to extract the knowledge that people are interested in, that is, correct, unknown, potentially valuable, and can be received and utilized by users from a large amount of data. Data mining has been widely used in various fields, more and more scholars begin to pay attention to the field of data mining. With the popularization of data mining related knowledge in more and more fields, the practical tools of data mining have been paid more and more attention. The research focus of data mining is gradually shifting to the tool of data mining system, and more attention is paid to the integration of various discovery strategies and technologies. Data mining tools are mainly concentrated in some traditional industries, such as banking, manufacturing, sales and other industries facing transformation, data mining tools are constantly produced and improved. Based on the above scenarios and requirements, this paper studies and designs a data mining modeling tool based on process. The tool provides a process modeling method based on web services, which aims to provide a unified modeling platform for process developers, reduce the workload of developers and meet the dynamic needs of different developers. With the help of web services, the modeling method of process-based data mining can meet the task requirements of developers by assembling third-party services into a new process. The advantage of this method is that the developer does not need to know too much about data mining, and completely hides the details of the implementation of the underlying data mining services, and only needs to call the related web services based on data mining. The data mining process can be constructed by calling the interface provided by the service and assembling these services. This tool can significantly improve the modeling efficiency, and can greatly reduce the original IT support cost for IT personnel, and improve the scheduling and data operation efficiency. Developers can customize the details of each step of the modeling process and track the execution of the process in real time. In order to further reduce the workload of developers, the system provides developers with a service library for process building. The tasks of data preprocessing, mining algorithm and algorithm evaluation are encapsulated in the process of data mining, which is provided to the developer in the form of service template. With the help of the service library provided by the system, the developer can complete the modeling process from data input, data splicing, feature extraction, model training, model evaluation, model application, and significantly improve the efficiency of process modeling. When setting up the data mining process, the developer can choose to import the third-party service or call the system to provide the service, which greatly increases the flexibility of the process development. In this paper, the background and significance of data mining modeling system are introduced, and the related technologies and fields are studied. Then, the functional requirements of the data mining modeling system are analyzed from the overall goal of this paper. According to the requirements, the overall architecture of the system and the interaction interface between modules are designed, and the implementation of the core modules is introduced in detail. Finally, the system is tested to verify the rationality of the system.
【學(xué)位授予單位】：北京郵電大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2016
【分類號】：TP311.13;TP311.52

【參考文獻】

相關(guān)期刊論文 前10條

1 汪寶彬;汪玉霞;;隨機梯度下降法的一些性質(zhì)(英文)[J];數(shù)學(xué)雜志;2011年06期

2 樊瑋;匡載華;;基于UML的業(yè)務(wù)流程建模及到BPEL4WS的轉(zhuǎn)換[J];航空計算技術(shù);2010年01期

3 王雷;徐立臻;;BPEL建模工具中業(yè)務(wù)流程模型到BPEL程序的轉(zhuǎn)換算法[J];計算機與數(shù)字工程;2009年09期

4 姚世軍;謝蕾;吳之鐵;;基于WS-BPEL的服務(wù)組合流程建模工具的設(shè)計與實現(xiàn)[J];計算機工程與設(shè)計;2009年17期

5 倪晚成;劉連臣;吳澄;;Web服務(wù)組合方法綜述[J];計算機工程;2008年04期

6 明德廷;李娟;尹怡欣;;基于Web Service的數(shù)據(jù)挖掘系統(tǒng)的研究與應(yīng)用[J];計算機工程與設(shè)計;2007年09期

7 郭黎明;張艷珍;;A Distributed Data Mining System Based on Multi-agent Technology[J];Journal of DongHua University;2006年06期

8 李景霞,侯紫峰;Web服務(wù)組合綜述[J];計算機應(yīng)用研究;2005年12期

9 王冠,司建輝,楊昌鋒;數(shù)據(jù)挖掘系統(tǒng)研究[J];北京工業(yè)大學(xué)學(xué)報;2005年04期

10 汪加才,江效堯;一個開放數(shù)據(jù)挖掘系統(tǒng)的分析與設(shè)計[J];計算機工程與應(yīng)用;2004年24期

相關(guān)碩士學(xué)位論文 前10條

1 王華;基于YARN的數(shù)據(jù)挖掘系統(tǒng)的設(shè)計與實現(xiàn)[D];北京郵電大學(xué);2015年

2 陳夢飛;A銀行嵌入式數(shù)據(jù)挖掘的應(yīng)用研究[D];北京交通大學(xué);2015年

3 榮偉正;基于web的數(shù)據(jù)挖掘系統(tǒng)設(shè)計與實現(xiàn)[D];湖南大學(xué);2014年

4 蘇曉旭;Zmining數(shù)據(jù)挖掘系統(tǒng)中數(shù)據(jù)結(jié)構(gòu)的設(shè)計與實現(xiàn)[D];鄭州大學(xué);2014年

5 張軻智;基于web的數(shù)據(jù)挖掘系統(tǒng)設(shè)計與實現(xiàn)[D];電子科技大學(xué);2013年

6 楊金花;基于改進預(yù)測模型的數(shù)據(jù)挖掘系統(tǒng)研究與實現(xiàn)[D];電子科技大學(xué);2013年

7 張偉志;證券行業(yè)數(shù)據(jù)挖掘系統(tǒng)開發(fā)與設(shè)計[D];電子科技大學(xué);2013年

8 武桂云;基于hadoop平臺的分布式數(shù)據(jù)挖掘系統(tǒng)研究與設(shè)計[D];天津大學(xué);2012年

9 張穎;有關(guān)共軛梯度法的一些研究[D];大連理工大學(xué);2012年

10 楊敏煜;基于改進聚類算法的數(shù)據(jù)挖掘系統(tǒng)的研究與實現(xiàn)[D];電子科技大學(xué);2012年

，

本文編號：2307283