Computational Web Intelligence for Wired and Wireless

Computational Web Intelligence for Wired and Wireless Applications Yan-Qing Zhang Department of Computer Science Georgia State University Atlanta, GA 30302 -4110 yzhang@cs. gsu. edu

Outline n n n Introduction Computational Intelligence Web Technology Computational Web Intelligence (CWI) Wired and Wireless Applications Conclusion and Future Work 2

Introduction Qo. I (Quality of Intelligence) of e-Business n WI = AI + IT WI (Web Intelligence) exploits Artificial Intelligence (AI) and advanced Information Technology (IT) on the Web and Internet. (Zhong, Liu, Yao and Ohsuga) at Proc. the 24 th IEEE Computer Society International Computer Software and Applications Conference (COMPSAC 2000), n 3

Introduction (cont. ) “CI is a subset of AI”, n “CI is not a subset of AI, there is an overlap between AI and CI”. n In general, CI AI. crisp logic and rules in AI, and fuzzy logic and rules in CI (Zadeh). n Motivation: “Input CI onto Web? ” n 4

Computational Intelligence n n n n fuzzy computing (FC) neural computing (NC), evolutionary computing (EC), probabilistic computing (PC), granular computing (Gr. C) rough computing (RC). … 5

Web Technology a hybrid technology including computer networks, the Internet, wireless networks, databases, search engines, client-server, programming languages, Web-based software, security, agents, e-business systems, and other relevant techniques. 6

Computational Web Intelligence (Zhang and Lin, 2002) n Uncertainty on the Web (FLINT 2001 at BISC at UC Berkeley http: //www-bisc. cs. berkeley. edu/) (Zhang, et al, 2001 (a), (b) (c)) n CWI = CI + WT (Zhang and Lin, 2002) CWI is a hybrid technology of Computational Intelligence (CI) and Web Technology (WT) on wired and wireless networks. CWI is dedicating to increasing Qo. I of e-Business applications with uncertain data on the Internet and wireless networks. 7

Computational Web Intelligence (cont. ) (Zhang and Lin 2002) n n n n Fuzzy Web Intelligence Neural Web Intelligence Evolutionary Web Intelligence Probabilistic Web Intelligence Granular Web Intelligence Rough Web Intelligence Hybrid Web Intelligence 8

9

n n n Preface. . . v Introduction to Computational Web Intelligence and Hybrid Web Intelligence. . . . xviii Part I: Fuzzy Web Intelligence, Rough Web Intelligence and Probabilistic Web Intelligence. . . 1 Chapter 1. Recommender Systems Based on Representations. . . 3 Chapter 2. Web Intelligence: Concept-based Web Search. . . . 19 Chapter 3. A Fuzzy Logic Approach to Answer Retrieval from the World -Wide-Web. . . . 53 Chapter 4. Fuzzy Inference Based Server Selection in Content Distribution Networks. . . . 77 Chapter 5. Recommendation Based on Personal Preference. . . …. . 101 Chapter 6. Fuzzy Clustering and Intelligent Search for a Web-based Fabric Database. . . . . 117 Chapter 7. Web Usage Mining: Comparison of Conventional, Fuzzy and Rough Set Clustering. . . . 133 Chapter 8. Towards Web Search Using Contextual Probabilistic Independencies. . . . 149 10

n n n n n Part II: Neural Web Intelligence, Evolutionary Web Intelligence and Granular Web Intelligence 167 Chapter 9. Neural Expert System for Vehicle Fault Diagnosis via The WWW. . . . . 169 Chapter 10. Dynamic Documents in The Wired World. . 183 Chapter 11. Proximity-based Supervision for Flexible Web Page Categorization. . . . 205 Chapter 12. Web Usage Mining: Business Intelligence From Web Logs. . 229 Chapter 13. Intelligent Content-Based Audio Classification and Retrieval for Web Application. . . . 257 11

n n n n Part III: Hybrid Web Intelligence and e-Applications 283 Chapter 14. Developing an Intelligent Multi-Regional Chinese Medical Portal. . . . . 285 Chapter 15. Multiplicative Adaptive User Preference Retrieval and Its Applications to Web Search. . . . 303 Chapter 16. Scalable Learning Method to Extract Biological Information from Huge Online Biomedical Literature. . . . . 329 Chapter 17. i. MASS: An Intelligent Multi-resolution Agent-based Surveillance System. . . . 347 Chapter 18. Networking Support for Neural Network-based Web Monitoring and Filtering. . . . 369 Chapter 19. Web Intelligence: Web-based BISC Decision Support System (WBICS-DSS). . . . 391 Chapter 20. Content and Link Structure Analysis for Searching the Web. 431 Chapter 21. Mobile Agent Technology for Web Applications. . 453 Chapter 22. Intelligent Virtual Agents and the WEB. . . 481 Chapter 23. Data Mining in Network Security. . . . 501 Chapter 24. Agent-supported WI Infrastructure: Case Studies in Peer-topeer Networks. . . . . 515 Chapter 25. Intelligent Technology for Content Monitoring on the Web. . 539 12

Wired and Wireless Applications CWI has various applications in intelligent e-Business on the Internet and on wireless mobile networks. 1. Neural-Net-based online Stock Agents, 2. Personalized Mobile Phone Agents, 3. Mobile Wireless Shopping Agents, 4. Mobile Wireless Fleet Application (Yamacraw Research Project). 13

Fuzzy Neural Web Agents for Stock Prediction (Zhang, et al, 2001) To implement this stock prediction system, Java Servlets, Java Script and Jdbc are used. SQL is used as the back-end database. Data file Java conversion program SQL table

Fig 1. Graph of Predicted and Real values for dow stock using complete data (Zhang, et al, 2001)

Personalized Wireless Information Agents for Mobile Phones

Personalized Weather Agent

Ø Mobile Wireless Shopping Agents Local go Fuzzy user Ranking Display search result Search Agent message dispatch generate Local Agent time out counter=2 counter=1 Search Agent store 2 go with result Search Agent File go Search Agent message search result Local File with result store 1 go go 18

Mobile Fleet Application (Yamacraw Research Project) Web and Data Center User n n n Depot 1 Depot 2 n Automated scheduling of pickups and deliveries Distributed design Emergency Handling: On-the-fly scheduling of package exchanges between trucks (rendezvous – peer-topeer interaction) Demo 19

20

Truck to Truck Communication • Sy. D Listen er listene r Sy. D Engine TDB Truck 1 Truck App. O Sy. D Engine DBS: Database service TDB: Truck database • A selected (Truck 2) peer resolves the request using Its own Sy. D Listener and Engine. • Sends the result back to the calling peer (Truck 1). • Truck App. O A truck (Truck 1) sends a request to the Sy. D Listener on a peer truck using Sy. D Engine “invoke” method. IP address of peers are resolved using the Sy. D directory service running in a central location • Each device is capable of functioning as client or server. TDB Truck 2 21

Conclusion CWI based on CI and WT, a new research area, is proposed to increase the Qo. I of e. Business applications. CWI has a lot of wired and wireless applications in intelligent e-Business. FWI, NWI, EWI, PWI, GWI, RWI, and HWI are major CWI techniques currently. 22

Future Work n n n CWI on wired and mobile wireless networks. Web Data Mining and Knowledge Discovery. Intelligent wireless mobile PDAs (do smart e. Business, Homeland Security, etc. ) Intelligent Wireless Mobile Agents (in cars, houses, offices, etc. ) Intelligent Bioinformatics on the Web CWI and Grid Computing. 23

References [1] Y. -Q. Zhang, A. Kandel, T. Y. Lin and Y. Y. Yao (eds. ), “Computational Web Intelligence: Intelligent Technology for Web Applications, ” Series in Machine Perception and Artificial Intelligence, volume 58, World Scientific, 2004. [2] Y. -Q. Zhang and T. Y. Lin, “Computational Web Intelligence (CWI): Synergy of Computational Intelligence and Web Technology, ” Proc. of FUZZ-IEEE 2002 of World Congress on Computational Intelligence 2002: Special Session on Computational Web Intelligence, pp. 1104 -1107, Honolulu, May 2002. [3] M. Atlas and Y. -Q. Zhang, “Fuzzy Neural Web Agents for Efficient NBA Scouting, ” Web Intelligence and Agent Systems: An International Journal, vol. 6, no. 1, pp. 83 -91, 2008. [4] Y. -Q. Zhang, S. Hang, T. Y. Lin and Y. Y. Yao, “Granular Fuzzy Web Search Agents, ” Proc. of FLINT 2001, pp. 95 -100, UC Berkeley, Aug. 14 -18, 2001. [5] Y. -Q. Zhang, S. Akkaladevi, G. Vachtsevanos and T. Y. Lin, “Fuzzy Neural Web Agents for Stock Prediction, ” Proc. of FLINT 2001, pp. 101 -105, UC Berkeley, Aug. 14 -18, 2001. [6] Y. Tang and Y. -Q. Zhang, “Personalized Library Search Agents Using Data Mining Techniques, ” Proc. of FLINT 2001, pp. 119 -124, UC Berkeley, Aug. 14 -18, 2001. 24

Thank you! n Any Question? 25