面向商业智能的数据挖掘算法和多智能体系统的体系结构以及优化

面向商业智能的数据挖掘算法和多智能体系统的体系结构以及优化

论文摘要

目前,对于组织机构,数据挖掘大势所趋、广为应用。因为不管何时何地,一切皆为数据。为了有效应对,他们聘请数据处理专家,探索动态的开发工具和方法。热衷于隐私保护的人则关心数据的处理、管理以及再应用。科学家、数据员,技术专家以及企业家期望找到新的方法来搜索、提取,并预测可用的信息和知识。数据挖掘,商业智能和多智能体系统作为一种单一的搜索技术已经非常成熟。在当代商务领域,科技及工程理论、数据建构、存储设施都已基本完备。然而用工具和专属技术来挖掘有价值的信息,准确获取和传递数据,仍面临挑战。因此,发展一种信息整合的、智能的方法处理、管理和有效地利用数据,涌现出大量的需求,也是迎接数据时代挑战的根本之路。在信息技术发展,商业的动态性,用户的需求以及导致商业成功的背景中,对于优化工具、智能体的应用及其性能,至关重要。然而,单一的工具或是代理技术不能满足目前的商业需求,我们需要寻求一种可视化的、在大数据中挖掘有效信息的途径,在此前提下,多方法融合的智能的数据处理、管理和使用方法应运而生。本文研究了一种一般化的方法,对于发展物理、生物以及社会系统的一般模型有贡献意义。本研究主要侧重于研究大数据是如何有效利用的,来更好地发挥出信息的力量。在本文中,我们引入了基于商业智能的数据挖掘、多智能体系统的集成模式,提出了一种简单的一般推测模型和挖掘算法。该方法可以可扩展性地发展多个商业智能工具,可以让用户方便地根据所需传播信息的动态架构。为了搜索大范围的数据,我们采用了一种系统新颖的方法,优化了现有的工具和应用智能体的技术。为现代商业系统的优化提供了一种可行方法。这些尝试是对传统挖掘技术的革新,可以生成一种通用模型,结合了物理、生物、社会系统的特点,进行商业智能系统的配置。此外,多种挖掘算法和技术也经过了作者深入的比较、探讨和改进。此外,本文还尝试了两种不同分支间的重要交融。一是商业和科学的交融,把传统的挖掘方法和商务智能相结合,主要体现在第3、5、6和7章。技术细节和算法可以通过商务智能工具DW,OLAP, OLTP和在其他动态建模基础上量化和实现,并允许用户按需接入和传播数据。这是一种对于高级工具使用的代理技术在大数据上应用的新尝试,可以用于优化现有的商务智能系统。第二点融合是应用工具与代理技术的交融,并以可视化的形式呈献给用户,这要求决策者、用户、研究人员在传统商务流程中加入软实力来重建商务模式,第8、9和10章做了具体探讨。全文第4章是商务预期探讨,而第11章是对现有方法的融合及尝试。因此,没有工具的应用和代理技术,就谈不上商务决策的优化和效率。数据挖掘和多智能体系统的集成是处理大数据的可视化工具。这些可视化信息对于商业或者整个组织机构有易读性,为普适计算,数据数字化提供了一条有效路径。目前,对于商业组织,数字化已经不再是一个选择,而成为必备的技术手段,因为集成和智能方法的问题求解,性能优化和风险削减,也是现代商务的基本任务和方法,即“从信息时代到知识时代的转变”,是研究集成的应用和代理技术的科学方法。集成的应用和代理技术能够动态的自适应的解决各种问题。归根结底,数据挖掘是以解决实际问题为导向的。本论文的重要成果总结如下:-处理商业挑战的单一方法在诸多原因的限制下并不都是有效的。在本文中,我们讨论并提出了一种新颖的智能工具和智能体集成技术。该技术对于特定的商业需求,能给出最好的自适应方案,供决策者做出最优的判断。-可视化图示对于情景的概念化具有挑战。纵观文献,人们尝试了各种模型来应对这些棘手的问题。本文提出并发展了一种通用商业智能模型,并集成了数据挖掘算法。这些挖掘算法对于挖掘复杂而海量的数据是非常有必要的,并且它可以通过描画许多相关知识来辅助商业智能系统以及信息挖掘过程,以扩展用户的能力。-该研究所用的通用模块提供了一个共用框架,对概念和词表进行精确的识别、描绘,在分析各种商业活动中,缩短了决策者和专家之间的距离,可以对商务问题多角度地进行描述。本论文工作将分为三大部分,包括:(i)第1,2,3章讨论了研究的理论与技术,并做出需求分析。(ii)第4,5,6,7章引入模型与算法,进行新的系统构想。(iii)第8,9,10,11章提供了建模和性能评价,并展开了具体应用,第12章是结论与未来的研究方向。本论文的第6章到第11章由12篇论文作为支撑,其中包括6篇已发表文章(EI:2,国际会议:2,数据挖掘方面的书籍章节:2)。其他的6篇文章(其中1篇已录用)在审稿修订中。除此之外,本研究工作由9个相关主题课堂作业支撑,其中5篇论文已被国际会议收录。

论文目录

  • ABSTRACT
  • 摘要
  • ACKNOWLEDGEMENTS
  • TABLE OF CONTENTS
  • LIST OF FIGURES
  • LIST OF TABLES
  • LIST OF ACRONYMS
  • PART I: RESEARCH FOUNDATION, ART OF TECHNOLOGIES AND REQUIRMENT ANALYSIS
  • CHAPTER 1: INTRODUCTION
  • 1.1 MOTIVATION
  • 1.2 RESEARCH GOALS AND OBJECTIVES
  • 1.3 THE RESEARCH ISSUES
  • 1.4 RESEARCH METHODOLOGY
  • 1.5 SCOPE OF THE DISSERTATION
  • 1.6 TERMINOLOGY AND NOTIONS
  • 1.6.1 Fundamental terms
  • 1.6.2 Notions
  • 1.7 CONTRIBUTIONS OF DISSERTATION
  • 1.8 ORGANIZATION OF DISSERTATION
  • 1.9 THE PHILOSOPHY AND BASIC CONCEPTS OF INTEGRATION OF THE INTELLIGENT BUSINESS PROCESS
  • 1.9.1 Data mining
  • 1.9.1.1 Data mining blessing
  • 1.9.1.2 Data mining disambiguation
  • 1.9.1.3 Data mining is a key to integrated technology
  • 1.9.2 Agents and multi‐agent systems
  • 1.9.3 Business intelligence
  • 1.10 COMPLEXITY OF INTEGRATION BUSINESS INTELLIGENCE MODEL
  • 1.11 SUMMARY
  • CHAPTER 2: RELATED WORK AND ART OF TECHNOLOGIES
  • 2.1 BUSINESS INTELLIGENCE EXPECTATIONS AND CHALLENGES
  • 2.2 INTEGRATED APPROACH TOWARDS BUSINESS PROCESSING
  • 2.2.1 Tools integration framework
  • 2.2.1.1 Data capture/acquisition
  • 2.2.1.2 Data storage
  • 2.2.1.3 Data access and analysis
  • 2.2.1.4 Data warehouse, database and OLAP
  • 2.2.1.5 Mining techniques and algorithms
  • 2.2.2 Agents as a mining tool
  • 2.2.3 Agent requirements
  • 2.3 ARCHITECTURE FOR MODERN BUSINESS ONTOLOGICAL INTEGRATING
  • 2.3.1 Integrating of data mining ontology
  • 2.3.2 Integrating of agent (multi agent system)s ontology
  • 2.3.3 Integrating of business intelligence ontology
  • 2.4 DATA INTEGRATION AND DATA WAREHOUSING
  • 2.5 DATA MINING ALGORITHMS AND BUSINESS INTELLIGENCE MODELING
  • 2.5.1 Mining algorithms for business intelligence modeling
  • 2.5.2 The demand of integrating for data mining with business intelligence
  • 2.5.3 Mining structure for business intelligent modeling
  • 2.6 AN INTEGRATING FOR DATA MINING AND BUSINESS INTELLIGENCE
  • 2.6.1 The need of Integrating for data mining and multi agent systems
  • 2.6.2 Integrated application
  • 2.6.3 Integration visualization
  • 2.7 AGENTS AND MULTI-AGENT SYSTEMS
  • 2.7.1 Agents
  • 2.7.2 Multi agent systems
  • 2.7.3 Foundation for intelligent physical agents
  • 2.7.4 Multi agent system development platform
  • 2.8 MULTI AGENT DATA MINING
  • 2.8.1 Central learning strategy
  • 2.8.2 Meta-learning strategy
  • 2.8.3 Hybrid-learning strategy
  • 2.8.4 Generic multi-agent data mining
  • 2.9 COMPLEXITIES IN INTEGRATING A PROCESS
  • 2.10 SUMMARY
  • CHAPTER 3: INTEGRATION CONCEPTS, FUNDAMENTALS, APPROACH AND REQUIREMENT ANALYSIS
  • 3.1 THE FUNDAMENTAL OF INTEGRATION CONCEPTS AND METHODOLOGY
  • 3.1.1 Integration for mining tools fundamental and concepts
  • 3.1.1.1 Data mining driven agent integration
  • 3.1.1.2 Agent driven data exploration and tool's integration
  • 3.1.1.3 Business intelligent driven business packages and agent integration
  • 3.1.1.4 Generic model concepts and fundamental integration
  • 3.1.2 Concepts of mining tools and agent's integration
  • 3.2 INTEGRATED MODELING REQUIREMENT ANALYSIS AND DESIGN
  • 3.2.1 Structural Analysis and design
  • 3.2.2 Models operational analysis and integration
  • 3.3.3 Conceptual measuring and refining of integrated modeling
  • 3.3 INTEGRATING FOR MINING ALGORITHMS
  • 3.3.1 Choosing the right mining algorithm
  • 3.3.2 Concepts and methods of integrated data architecture
  • 3.4 FUNDAMENTAL TO INTEGRATING FOR TOOLS AND AGENTS
  • 3.4.1 Application tools driven agent integrations
  • 3.4.2 Agent driven application tool's integrations
  • 3.5 INTEGRATION PROTOCOL
  • 3.5.1 Data mining agent protocol
  • 3.5.2 Agent registration protocol
  • 3.6 THE ARCHITECTURE OF INTEGRATION
  • 3.6.1 Integrated modeling algorithms
  • 3.6.2 Modeling behavior protocol as of integrated business intelligence
  • 3.6.3 Equations and model applications
  • 3.6.3.1 Model types
  • 3.6.3.2 Model complexity
  • 3.7 DATA MINING PREDICTIVE AND DESCRIPTIVE MODELING METHODS
  • 3.7.1 Predictive modeling
  • 3.8.1.1 Information integration tasks
  • 3.7.1.2 Information integration model
  • 3.7.1.3 Building predictive model
  • 3.7.2 Descriptive modeling
  • 3.8 SUMMARY
  • PART II: MODELING DEVELOPMENT AND ITS ANALYTICS
  • CHAPTER 4: BUSINESS INTELLIGENCE DATA MINING AND DATA WAREHOUSE
  • 4.1 INTEGRATION PROCESS IN THE DATA WAREHOUSE
  • 4.1.1 Data Warehouse Architecture
  • 4.1.2 Data warehouse and decision support systems
  • 4.1.3 Data warehouse data preparation
  • 4.2 BUSINESS INTELLIGENCE DATA MINING OF DATA WAREHOUSING MODELS
  • 4.2.1 Data warehouse modeling techniques
  • 4.2.2 Online transaction processing
  • 4.2.3 Online analytical processing
  • 4.2.4 Online transaction processing Vs online analytical processing
  • 4.3 THE PARADIGM OF DATA WAREHOUSE INTEROPERABILITY AND APPLICATION
  • 4.3.1 Features of and applications
  • 4.3.2 Star Schema
  • 4.3.3 Snowflake schema
  • 4.3.4 Dimension reduction
  • 4.3.5.1 Dimension table and fact table
  • 4.3.5.2 Correlation analysis
  • 4.3.5.5 Principal component analysis
  • 4.3.6 Data granularity
  • 4.4 SUMMARY
  • CHAPTER 5: ARCHITECTURE OF INTEGRATING EMERGING TECHNOLOGY
  • 5.1 ARCHITECTURE CONCEPT AND METHODS FOR INTEGRATION
  • 5.1.1 The concepts of enable technology
  • 5.1.1.1 Elements of service oriented architecture
  • 5.1.1.2 Gird service architecture
  • 5.1.2 Integration based application service architecture
  • 5.1.3 Role based integrated model platform architecture
  • 5.1.3.1 Service location
  • 5.1.3.2 Service instantiation
  • 5.1.3.3 Task based instantiation
  • 5.2 FUNDAMENTAL OF INTEGRATED EMERGING TECHNOLOGY
  • 5.3 THE BASIC PRIMITIVES OF EMERGING FOR BUSINESS INTELLIGENCE DATA MINING
  • 5.3.1 Distributed integration architecture
  • 5.3.2 Concepts and mechanisms of integration
  • 5.3.3 Data centric integration mechanism
  • 5.4 DATA MINING APPLICATION AND TRENDS FOR EMERGING TECHNOLOGY
  • 5.5 PRINCIPLES OF DIMENSIONAL DATA MODELING
  • 5.6 PREDICTION BUSINESS MODELS
  • 5.6.1 Predictive/Supervised Business Modeling
  • 5.6.1.1 Supervised structure prediction
  • 5.6.1.2 Concept and fundamental to decision trees based predictive modeling
  • 5.6.1.3 Algorithmic framework decision tree based predictive modeling
  • 5.6.1.4 Fundamental of tree pruning for predictive modeling
  • 5.6.1.5 Concepts of minimum description length based pruning
  • 5.6.2 Predictive business model methodology and application
  • 5.6.2.1 Logistic regression model
  • 5.6.2.2 Naive Bayes approach text classification
  • 5.6.3 Unsupervised predictive business model
  • 5.6.3.1 Reduction for unsupervised to supervised
  • 5.6.3.2 Unsupervised algorithms
  • 5.6.3.3 Rule based unsupervised prediction modeling
  • 5.6.4 Clustering
  • 5.8 SUMMARY
  • CHAPTER 6: DEVELOPMENT OF GENERIC BUSINESS INTELLIGENCE MODEL
  • 6.1 PARADIGM OF GENERIC MODEL
  • 6.1.1 Model integration platform and information portal
  • 6.1.2 Information access and distributions
  • 6.1.3 Data exploratory and analysis
  • 6.2 REQUIREMENT ANALYSIS AND DEVELOPMENT OF BUSINESS INTELLIGENCE GENERIC MODEL
  • 6.2.1 A framework for integration of data mining and agents with business intelligence
  • 6.2.1.1 The proposed generic architecture of a data mining framework
  • 6.2.1.2 Generic data integration model
  • 6.2.2 Model complexities in integrating process
  • 6.2.3 Conceptual of theorizing and modeling
  • 6.3 DATA INTEGRATIONS AND MINING ALGORITHMS
  • 6.3.1 System development
  • 6.3.2 Data mart design
  • 6.4 MINING TECHNIQUES, ALGORITHMS AND MODELING FOR GENERIC BUSINESS INTELLIGENCE
  • 6.4.1 The techniques of decision tree
  • 6.4.2 Association Rules
  • 6.4.2.1 Frequent pattern mining
  • 6.4.2.2 A priori algorithms
  • 6.4.3 CLUSTERING ANALYSIS
  • 6.4.3.1 Clustering algorithms
  • 6.4.3.2 Similarity measures
  • 6.5 MODEL DEVELOPMENT (BUSINESS INTELLIGENCE) LIFE CYCLE
  • 6.6 SUMMARY
  • CHAPTER 7: ANALYTIC OF INTEGRATION FOR TOOLS AND AGENTS TOWARDS GENERIC MODEL
  • 7.1 DATA INTEGRATION AND PREPARATION
  • 7.1.1 Data type and design
  • 7.1.2 Data processing and flow
  • 7.1.3 A process of knowledge discovery
  • 7.2 THE PARADIGM OF INFORMATION INTEGRATION
  • 7.2.1 Issues of information integration
  • 7.2.2 Information integration extending the data warehousing
  • 7.2.3 Information integration for performance and scalability of business paradigm
  • 7.3 PREDICTIVE ANALYTICS AND MINING PROCESS: STRATEGIC IMPLEMENTATION
  • 7.3.1 Big Picture: CRISP-DM based integrated BIDM
  • 7.3.2 Integrated BI modeling based knowledgy discovery
  • 7.4 TECHNIQUES FOR EXTRACTION OF DATA
  • 7.4.1 Extract, Transform and Load (ETL)
  • 7.4.2 Business Intelligence Data Storage and management
  • 7.4.3 Business process intelligence (BPI)
  • 7.5 DATA TRANSFORMATION DESIGN
  • 7.6 DATA STAGING AND QUALITY
  • 7.7 DATA VISUALIZATION
  • 7.8 SUMMARY
  • PART III: MODEL PERFORMANCE EVALUATION AND APPLICATION
  • CHAPTER 8: DATA EXPLORATION AND EVALUATION PERFORMANCE FOR GENERIC MODELING OPTIMIZATION199
  • 8.1 DATA EXPLORATION TOWARDS GENERIC MODEL PERFORMANCE OPTIMIZATION
  • 8.1.1 Distributed databases systems
  • 8.1.2 Searching and exploration
  • 8.1.2.1 Search in generic model performance
  • 8.1.2.2 Exploration in design
  • 8.2 COMPONENT BASED GENERIC MODEL EXPLORATION
  • 8.2.1 Generic model performance (degree of freedom)
  • 8.2.1.1 Integrated (meta models) performance measure
  • 8.2.1.2 Performance optimization and formulation
  • 8.2.1.3 Model personalization and customization
  • 8.2.2 Generic model framework and performance optimization
  • 8.2.3 Generic model architectural view as performance measurement
  • 8.3 GENERIC MODEL APPLICATION AND OPTIMIZATION
  • 8.4 MODEL PERFORMANCE EVALUATION AND EXPLORATION
  • 8.4.1 A Modeling performance and exploration framwork
  • 8.4.2 Model performance evaluation and optimization
  • 8.5 GENERIC MODEL PERFORMANCE TESTING AND VALIDATION
  • 8.5.1 View for architectural performance validation
  • 8.5.2 Input-output modeling based test for generic model
  • 8.5.3 Accuracy modeling validation process
  • 8.6 DATA MINING AND AGENT BASED MODELING EXPLORATION
  • 8.6.1 Data mining based agent models exploration
  • 8.6.2 Agent based modeling exploration
  • 8.7 FUZZY LOGIC AND GENETIC ALGORITHM BASED EXPLORATION
  • 8.7.1 Fuzzy logic and data mining
  • 8.7.2 Genetic algorithms
  • 8.7.3 Fuzzy-genetic algorithms integrating
  • 8.8 SUMMARY
  • CHAPTER 9: GENERIC MODEL PERFORMANCE AND EVALUATIONS: DATA MINING AND BUSINESS INTELLIGENCE APPLICATIONS
  • 9.1 INFORMATION REQUIREMENTS FOR BUSINESS SUCCESS
  • 9.1.1 Quality of integrating for data mining and business intelligence
  • 9.1.1.1 Data mining on what kind of data?
  • 9.1.1.2 Data mining tool does data scoring
  • 9.1.1.3 Business intelligence tool does data scoring
  • 9.1.1.4 Paradigm of data mining with business intelligent system architecture
  • 9.1.2 Information value and an application
  • 9.1.2.1 Market basket analysis
  • 9.1.2.2 Business fraud detection
  • 9.1.3 Information access and distribution system
  • 9.1.4 The benefit of integrating modeling of data mining in business intelligence
  • 9.1.5 A generic model for information quality assurance: Integrating approach
  • 9.2 DATA MINING: CONFLUENCE OF MULTIPLE TASKS AND APPLICATIONS
  • 9.3 KNOWLEDGE DISCOVERY IN INTEGRATED DATA
  • 9.3.1 Transforming data into information and knowledge
  • 9.3.2 Integrated modeling towards knowledge discovery
  • 9.3.3 Integrated knowledge discovery in databases
  • 9.4 INTEGRATED BUSINESS INTELLIGENCE APPLICATIONS AND PERFORMANCE
  • 9.5 HIGH PERFORMANCE AND PRIVACY PRESERVING DATA MINING
  • 9.5.1 Privacy of individual data
  • 9.5.2 Distrubuted data mining based privacy performance optimization
  • 9.6 SUMMARY
  • CHAPTER 10: THE PARADIGM OF MULTI AGENT SYSTEMS IN BUILDING INTEGRATED BUSINESS INTELLIGENCE
  • 10.1 FROM SINGLE APPLICATION INTO INTEGRATION
  • 10.1.1 Agents modeling requirements and applications
  • 10.1.1.1 Agents based modeling development
  • 10.1.1.2 Agents applied in business modeling applications
  • 101.1.3 Single agent Vs multi agent applications
  • 10.2 AGENT/MULTIAGENT BASED DECISION SUPPORT SYSTEM
  • 10.2.1 Integrated Multi agent framework
  • 10.2.2 Decision support system framework
  • 10.3 ONTOLOGY BASED INTEGRATION OF AGENTS AND DATA MINING
  • 10.3.1 Data mining based decision support system
  • 10.3.2 Data mining and agent based generic decision support system
  • 10.3.4 Paradigm decision support system techniques
  • 10.4 AGENT COMMUNICATIONS LANGUAGES
  • 10.4.1 Knowledge interchange format
  • 10.4.2 Knowledge query and manipulation format
  • 10.4.3 Foundation for intelligent physical agent (FIPA) communication language
  • 10.5 ONTOLOGY FOR INTEGRATION AGENT COMMUNICATION
  • 10.6 MULTI-AGENT SYSTEMS AND ITS APPLICATION IN BUSINESS INTELLIGENCE
  • 10.6.1 Software agents
  • 10.6.1.1 Integrated intelligent information agents
  • 10.6.1.2 Workflow management and virtual organizations as agents
  • 10.6.1.3 The essence of Software agent for business intelligence
  • 10.6.2 Applications for Agents with Physical or Virtual Bodies
  • 10.6.2.1 Autonomous Control Systems
  • 10.6.2.2 Traffic telemetric
  • 10.7 AGENT BASED BUSINESS INTELLIGENCE MODEL PERFORMANCE FRAMEWORK
  • 10.8 SUMMARY
  • CHAPTER 11: BUSINESS PROCESSING MANAGEMENT PERFORMANCE: SUCCESS OF BUSINESS INTELLIGENCE ...
  • 11.1 VISUALIZATION OF BUSINESS INTELLIGENT PERFORMANCE QUALITY
  • 11.1.1 Integrated business intelligence for business processing management performance
  • 11.1.2 Qualitative performance of business intelligence in business processing performance
  • 11.2 THE PARADIGM OF BUSINESS PROCESSING MANAGEMENT PERFORMANCE
  • 11.2.1 A paradigm of business process management workflow
  • 11.2.1.1 Adaptive workflow management
  • 11.2.1.2 Workflow process modeling
  • 11.2.1.3 Workflow engine and its interface
  • 11.2.2 Decision making performance
  • 11.2.2.1 Business knowledge creations
  • 11.2.2.2 Decision support system and knowledge management systems in diction making process
  • 11.2.3 Performance measurement systems as an entity
  • 11.3 ACHIEVING BUSINESS INTELLIGENCE IMPACT
  • 11.3.1 Integrating business intelligence with core business processes
  • 11.3.2 Choice of Techniques
  • 11.3.3 Measuring goal oriented business intelligence based business processing
  • 11.4 CRITICAL ISSUES IN BUSINESS INTELLIGENCE BASED BUSINESS PROCESSING
  • 11.4.1 Critical success factors
  • 11.4.2 Business processing value of business intelligence
  • 11.4.3 The reason of business intelligence for business processing
  • 11.5 BUSINESS PROCESS DESIGN, DEPLOYMENT AND ONGOING MAINTENANCE
  • 11.5.1 Business intelligence based business process design
  • 11.5.2 Business processing system deployment and implementation
  • 11.5.3 Business process maintenance
  • 11.6 CHANGE MANAGEMENT
  • 11.6.1 A business process change framework
  • 11.6.2 Performance measurement towards change management
  • 11.6.3 Business process model's re-engineering as change management
  • 11.7 SUMMARY
  • CHAPTER 12: CONCLUSION AND FUTURE RESEARCH DIRECTION
  • 12.1 CONCLUSION
  • 12.2 FUTURE RESEARCH DIRECTION
  • REFERENCES
  • 相关论文文献

    标签:;  ;  ;  ;  ;  ;  

    面向商业智能的数据挖掘算法和多智能体系统的体系结构以及优化
    下载Doc文档

    猜你喜欢