李肇林臺灣大學：資訊工程學研究所楊凱翔Yang, Kai-HsiangKai-HsiangYang2007-11-262018-07-052007-11-262018-07-052004http://ntur.lib.ntu.edu.tw//handle/246246/54117Today, we see widespread use of the mobile devices that have sufficient computing and networking capabilities to browse the World Wide Web (WWW). These network-enabled devices include a variety of types, such as cell phones, Personal Digital Assists (PDAs), Pocket PCs, handheld PCs, car navigation systems, and notebook PCs. However, one of the biggest challenges in the Web-browsing is the heterogeneity of the devices. All these devices have different capabilities in processors, memory, networking, screen sizes, input methods, and software libraries. Such heterogeneity has resulted in an inconvenience of Web-browsing so that multiple solutions are then needed for these devices. An inconsistency of different operation methods may further lower the user’s interest of using such devices. These problems can be divided into four primary problems according to their particularities and when they occur during the Web-browsing process. In this dissertation, we propose the Integrated Personal Proxy (IPP) system that contains four new functions for solving the corresponding four problems. First, the proposed system provides a personalized, versioned, and location- and browser-independent cache log management mechanism, for the users to search and manage their cache logs. Besides, one URL correction function is applied to search all possible URLs in the users’ cache logs for the incorrect URLs. In order to integrate the filtering and translation functions, one modular platform is designed in the proposed system to provide the users to determinate their required functions. Furthermore, a web object delivery (WOD) function is also introduced to deliver the web objects which can not be downloaded into the devices by means of the email or ftp methods. We had also evaluated the proposed system in the response time and disk space, and all the experimental results indicate that the proposed system is an efficient implementation. The ideas and the system design could be very beneficial to the future researches for these problems.TABLE OF CONTENTS II LIST OF TABLES V LIST OF FIGURES VI ABSTRACT VIII 1. INTRODUCTION 1 1.1. TERMINOLOGY 1 1.2. CURRENT PROBLEMS 2 1.2.1. The personal cache log management problem 4 1.2.2. The URL typographical error problem 5 1.2.3. The integration and customization for filtering and translation function problem 5 1.2.4. The web object downloading problem 6 1.3. THE PROPOSED SYSTEM 7 1.3.1. The consideration of system design 8 1.3.2. The personal cache log management (PCLM) function 9 1.3.3. The URL correction (URLC) function 9 1.3.4. The integration and customization for the filtering and translation (ICFT) function 10 1.3.5. The web object delivery (WOD) function 10 1.4. THE ORGANIZATION OF THE DISSERTATION 11 2. RELATED WORKS 12 2.1. THE PERSONAL CACHE LOG MANAGEMENT PROBLEM 12 2.2. THE URL TYPOGRAPHICAL ERROR PROBLEM 14 2.3. THE INTEGRATION AND CUSTOMIZATION FOR FILTERING AND TRANSLATION FUNCTION PROBLEM 16 2.4. THE WEB OBJECT DOWNLOADING PROBLEM 18 3. PERSONAL CACHE LOG MANAGEMENT FUNCTION 21 3.1. LOGICAL CONCEPTS 21 3.2. DESIGN CONSIDERATION 22 3.3. THE ARCHITECTURE OF THE PCLM COMPONENT 23 3.3.1. HTTP core element 24 3.3.2. Proxy authentication 24 3.3.3. URL parser 25 3.3.4. Object retrieval element 26 3.3.5. Cache 27 3.3.6. Repository 30 3.3.7. Cache log management functions 33 3.3.8. Batch dispatcher 34 3.3.9. Web object retriever 34 3.3.10. Cache Log Management Interface 34 3.4. LINK UPDATING METHOD 35 3.5. DISTRIBUTED STORAGE AND CENTRALIZED MANAGEMENT 37 3.6. PERFORMANCE EVALUATION 38 3.6.1. Implementation Environments 38 3.6.2. Evaluation Methodology 38 3.6.3. Experiment Setup 41 3.6.4. Experimental Results 45 3.7. SUMMARY 51 4. URL CORRECTION COMPONENT 52 4.1. LOGICAL CONCEPTS 52 4.2. DESIGN CONSIDERATION 53 4.3. BASIC CONCEPTS 54 4.3.1. Edit distance (the URL similarity measurement) 54 4.3.2. Algorithm to compute edit distance 54 4.3.3. N-grams 55 4.3.4. Number of the n-grams 56 4.3.5. Filter conditions 56 4.3.6. URL Correction Model 57 4.4. THE ARCHITECTURE OF THE URLC COMPONENT 58 4.4.1. Index Production 59 4.4.2. URL correction 63 4.4.3. User Interface 67 4.5. EXPERIMENTAL RESULTS 67 4.5.1. Testing Data Set 67 4.5.2. Implementation Environment 68 4.5.3. Experiment Results 69 4.6. SUMMARY 80 5. INTEGRATION AND CUSTOMIZATION FOR FILTERING AND TRANSLATION FUNCTION COMPONENT 81 5.1. THREE CATEGORIES OF MODULES 81 5.2. LOGICAL CONCEPTS 84 5.3. DESIGN CONSIDERATION 85 5.4. ARCHITECTURE OF ICFT COMPONENT 85 5.4.1. Module profile element 87 5.4.2. Module profile management element 89 5.4.3. User profile element 90 5.4.4. User profile management element 91 5.4.5. Request work flow decision 91 5.4.6. Translation Logging 93 5.5. EXPERIMENTAL RESULTS 94 5.6. SYSTEM ADVANTAGES 98 5.7. SUMMARY 99 6. WEB OBJECT DELIVERY COMPONENT 100 6.1. LOGICAL CONCEPTS 100 6.2. DESIGN CONSIDERATION 101 6.3. ARCHITECTURE OF THE WEB OBJECT DELIVERY COMPONENT 102 6.3.1. User profile element 103 6.3.2. Decision element 105 6.3.3. Process queue 105 6.3.4. Batch processor 106 6.3.5. Delivery programs 107 6.3.6. User Profile Management 107 6.4. IMPLEMENTATION AND EXPERIMENTAL RESULTS 107 6.4.1. System Environment 107 6.4.2. Experimental Results 107 6.5. SUMMARY 110 7. CONCLUSION AND FUTURE WORKS 111 REFERENCE 1133265326 bytesapplication/pdfen-USURL修正快取資料管理代理伺服器URL correctionproxycache log managementthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/54117/1/ntu-93-F86526004-1.pdf