融合相关数据.ppt

Fusing Data with Correlations Ravali Pochampally, Anish Das Sarma, Luna Dong, Alexandra Meliou, Divesh Srivastava subject,predicate,object subject,predicate,object subject,predicate,object subject,predicate,object subject,predicate,object subject,predicate,object subject,predicate,object subject,predicate,object subject,predicate,object subject,predicate,object subject,predicate,object subject,predicate,object subject,predicate,object subject,predicate,object subject,predicate,object subject,predicate,object subject,predicate,object subject,predicate,object subject,predicate,object subject,predicate,object subject,predicate,object subject,predicate,object Fusion in web data extraction extractor extractor extractor extractor extractor 1 Applications: Building knowledge bases, answer questions, facilitate data mining How can we purge wrong triples from the dataset? ? ? ? ? Knowledge triple S1 S2 S3 S4 Daniel Radcliffe, played role, Harry Potter ? ? ? Daniel Radcliffe, spouse, Bonnie Wright ? ? Daniel Radcliffe, acted in, Frankenstein ? Emma Watson, acted in, Harry Potter ? ? ? J. K. Rowling, acted in, Harry Potter ? ? ? Richard Harris, played role, Dumbledore ? Michael Gambon, played role, Dumbledore ? Tim Burton, directed, Harry Potter ? ? ? Daniel Craig, acted in, Harry Potter ? ? Rupert Grint, acted in, Harry Potter ? ? The data fusion problem correlated 2 Contribution: Fusion techniques that consider source quality and correlations anti-correlated bad good This talk 3 2 techniques High-level intuition 4 S1 S2 S3 Jagadish UMich ATT UMich Dewitt MSR MSR UWisc Bernstein MSR MSR MSR Carey UCI ATT BEA Franklin UCB UCB UMD Researcher affiliation source quality correlations evaluation future directions High-level intuition 5 S1 S2 S3 Jagadish UMich ATT UMich Dewitt MSR MSR UWisc Bernstein MSR MSR MSR Carey UCI ATT BEA Franklin UCB UCB UMD Researcher affiliation source quality correlations evaluation future directions Voting: Trust the majority High-level intuition 6