《data mining concepts and techniques second edition 数据挖掘概念与技术 第二版 韩家炜 第八章》由会员分享，可在线阅读，更多相关《data mining concepts and techniques second edition 数据挖掘概念与技术 第二版 韩家炜 第八章（70页珍藏版）》请在金锄头文库上搜索。
1、Data Mining: Concepts and Techniques Chapter 8 8.1. Mining data streams Jiawei Han and Micheline Kamber Department of Computer Science University of Illinois at Urbana-Champaign www.cs.uiuc.edu/hanj 2006 Jiawei Han and Micheline Kamber. All rights reserved. Date1Data Mining: Concepts and Techniques
2、Date2Data Mining: Concepts and Techniques Data and Information Systems (DAIS:) Course Structures at CS/UIUC nThree streams: Database, data mining and text information systems nDatabase Systems: nDatabase mgmt systems (CS411: Fall and Spring) nAdvanced database systems (CS511: Fall) nWeb information
3、systems (Kevin Chang) nInformation integration (An-Hai Doan) nData mining nIntro. to data mining (CS412: HanFall) nData mining: Principles and algorithms (CS512: HanSpring) nSeminar: Advanced Topics in Data mining (CS591HanFall and Spring) nText information systems and Bioinformatics nText informati
4、on system (CS410Zhai) nIntroduction to BioInformatics (CS598Sinha, CS498Zhai) Date3Data Mining: Concepts and Techniques Data Mining: Concepts and Techniques, 2ed. 2006 nSeven chapters (Chapters 1-7) are covered in the Fall semester nFour chapters (Chapters 8-11) are covered in the Spring semester Da
5、te4Data Mining: Concepts and Techniques Coverage of CS412UIUC (Intro. to Data Warehousing and Data Mining) n Introduction n Data Preprocessing n Data Warehouse and OLAP Technology: An Introduction n Advanced Data Cube Technology and Data Generalization n Mining Frequent Patterns, Association and Cor
6、relations n Classification and Prediction n Cluster Analysis Date5Data Mining: Concepts and Techniques Coverage of CS512UIUC (Data Mining: Principles and Algorithms) n Mining stream, time-series, and sequence data n Mining data streams n Mining time-series data n Mining sequence patterns in transact
7、ional databases n Mining sequence patterns in biological data n Graph mining, social network analysis, and multi-relational data mining n Graph mining n Social network analysis n Multi-relational data mining 10.Mining Object, Spatial, Multimedia, Text and Web data n Mining object data n Spatial and
8、spatiotemporal data mining n Multimedia data mining n Text mining n Web mining 11.Applications and trends of data mining n Data mining applications n Data mining products and research prototypes n Additional themes on data mining n Social impacts of data mining n Trends in data mining Date6Data Mini
9、ng: Concepts and Techniques Chapter 8. Mining Stream, Time -Series, and Sequence Data nMining data streams nMining time-series data nMining sequence patterns in transactional databases nMining sequence patterns in biological data Date7Data Mining: Concepts and Techniques Mining Data Streams nWhat is
10、 stream data? Why Stream Data Systems? nStream data management systems: Issues and solutions nStream data cube and multidimensional OLAP analysis nStream frequent pattern analysis nStream classification nStream cluster analysis nResearch issues Date8Data Mining: Concepts and Techniques Characteristi
11、cs of Data Streams nData Streams nData streamscontinuous, ordered, changing, fast, huge amount nTraditional DBMSdata stored in finite, persistent data setsdata sets nCharacteristics nHuge volumes of continuous data, possibly infinite nFast changing and requires fast, real-time response nData stream
12、captures nicely our data processing needs of today nRandom access is expensivesingle scan algorithm (can only have one look) nStore only the summary of the data seen thus far nMost stream data are at pretty low-level or multi-dimensional in nature, needs multi-level and multi-dimensional processing
13、Date9Data Mining: Concepts and Techniques Stream Data Applications nTelecommunication calling records nBusiness: credit card transaction flows nNetwork monitoring and traffic engineering nFinancial market: stock exchange nEngineering & industrial processes: power supply & manufacturing nSensor, moni
14、toring & surveillance: video streams, RFIDs nSecurity monitoring nWeb logs and Web page click streams nMassive data sets (even saved but random access is too expensive) Date10Data Mining: Concepts and Techniques DBMS versus DSMS nPersistent relations nOne-time queries nRandom access n“Unbounded” dis
15、k store nOnly current state matters nNo real-time services nRelatively low update rate nData at any granularity nAssume precise data nAccess plan determined by query processor, physical DB design nTransient streams nContinuous queries nSequential access nBounded main memory nHistorical data is impor
16、tant nReal-time requirements nPossibly multi-GB arrival rate nData at fine granularity nData stale/imprecise nUnpredictable/variable data arrival and characteristics Ack. From Motwanis PODS tutorial slides Date11Data Mining: Concepts and Techniques Mining Data Streams nWhat is stream data? Why Stream Data Systems? nStream data management systems: Issues and solutions nStream data cube and multidimensional OLAP analysis nS
