不连续跟不稳定数据管理英文版本.ppt

Contributors Ariel Fuxman, PhD Thesis Microsoft Search Labs Jim Gray SIGMOD 2008 Dissertation Award Periklis Andritsos, PhD Jiang Du, MS Elham Fazli, MS Diego Fuxman, Undergrad Limitations of Data Cleaning Semi-automatic process Requires highly-qualified domain experts Time consuming May not be possible to wait until the database is clean Operational systems answer queries assuming clean data Why is this Business Intelligence? Business intelligence (BI) refers to technologies, applications and practices for the collection, integration, analysis, and presentation of information. The goal of BI is to support better decision making, based on information. DBMS should provide meaningful query answers even over data that is dirty Uncertain Data Outline Introduction Semantics for dirty databases Contributions Conclusions * A Virtuous Cycle * Query Answering Data Integration Recognize and characterize inconsistent data Use knowledge about inconsistencies to: give better answers suggest ways to clean the database Beyond the Enterprise Can we apply principled models of inconsistency or uncertainty to the Web? Different assumptions Uncertainty in queries There’s never a “true” answer Challenge Build models based on user preferences Leverage massive repositories of user behavior data * THANK YOU Plug: Discovering Data Quality Rules, Fei Chiang Thursday 11:15am Research Session 33 * * * Existing approaches to this problem are based on data cleaning. * The problem of dirty data appears in this context even if the sources are clean… As a motivation, lets focus on a domain known in IT as CRM -- Customer Relationship Management. One of the goals of CRM is to integrate customer information from such disparate sources as ..... This domain is of interest to us because customer data is notoriously dirty and inconsistent. For merging, data integration tools provide much less support than for matching Matching and merging are two fundamental tasks in CRM Match