The practice of crowdsourcing is transforming the Web and giving rise to a new field.

@article{doan2011crowdsourcing, author = {Doan, Anhai and Ramakrishnan, Raghu and Halevy, Alon Y.}, title = {Crowdsourcing systems on the World-Wide Web}, journal = {Communications of the ACM}, publisher = {ACM}, address = {New York, NY, USA}, year = {2011}, volume = {54}, number = {4}, pages = {86--96}, url = {http://doi.acm.org/10.1145/1924421.1924442}, doi = {10.1145/1924421.1924442}, issn = {0001-0782}, keywords = {crowdsourcing, human, intelligence, social, computing, cirg, collective}, abstract = {The practice of crowdsourcing is transforming the Web and giving rise to a new field.} }

%0 = article %A = Doan, Anhai and Ramakrishnan, Raghu and Halevy, Alon Y. %C = New York, NY, USA %D = 2011 %I = ACM %T = Crowdsourcing systems on the World-Wide Web %U = http://doi.acm.org/10.1145/1924421.1924442

A primary challenge to large-scale data integration is creating semantic equivalences between elements from different data sources that correspond to the same real-world entity or concept. Dataspaces propose a pay-as-you-go approach: automated mechanisms such as schema matching and reference reconciliation provide initial correspondences, termed <i>candidate matches</i>, and then user feedback is used to incrementally confirm these matches. The key to this approach is to determine in what order to solicit user feedback for confirming candidate matches.</p> <p>In this paper, we develop a decision-theoretic framework for ordering candidate matches for user confirmation using the concept of the <i>value of perfect information (VPI)</i>. At the core of this concept is a <i>utility function</i> that quantifies the desirability of a given state; thus, we devise a utility function for dataspaces based on query result quality. We show in practice how to efficiently apply VPI in concert with this utility function to order user confirmations. A detailed experimental evaluation on both real and synthetic datasets shows that the ordering of user feedback produced by this VPI-based approach yields a dataspace with a significantly higher utility than a wide range of other ordering strategies. Finally, we outline the design of Roomba, a system that utilizes this decision-theoretic framework to guide a dataspace in soliciting user feedback in a pay-as-you-go manner.

@inproceedings{jeffery2008payasyougo, author = {Jeffery, Shawn R. and Franklin, Michael J. and Halevy, Alon Y.}, title = {Pay-as-you-go user feedback for dataspace systems}, booktitle = {Proceedings of the 2008 ACM SIGMOD international conference on Management of data}, publisher = {ACM}, address = {New York, NY, USA}, year = {2008}, pages = {847--860}, url = {http://doi.acm.org/10.1145/1376616.1376701}, doi = {10.1145/1376616.1376701}, isbn = {978-1-60558-102-6}, keywords = {matching, feedback, entity, schema, human, intelligence, semantic, social, computing, collective, linking, web}, abstract = {A primary challenge to large-scale data integration is creating semantic equivalences between elements from different data sources that correspond to the same real-world entity or concept. Dataspaces propose a pay-as-you-go approach: automated mechanisms such as schema matching and reference reconciliation provide initial correspondences, termed <i>candidate matches</i>, and then user feedback is used to incrementally confirm these matches. The key to this approach is to determine in what order to solicit user feedback for confirming candidate matches.</p> <p>In this paper, we develop a decision-theoretic framework for ordering candidate matches for user confirmation using the concept of the <i>value of perfect information (VPI)</i>. At the core of this concept is a <i>utility function</i> that quantifies the desirability of a given state; thus, we devise a utility function for dataspaces based on query result quality. We show in practice how to efficiently apply VPI in concert with this utility function to order user confirmations. A detailed experimental evaluation on both real and synthetic datasets shows that the ordering of user feedback produced by this VPI-based approach yields a dataspace with a significantly higher utility than a wide range of other ordering strategies. Finally, we outline the design of Roomba, a system that utilizes this decision-theoretic framework to guide a dataspace in soliciting user feedback in a pay-as-you-go manner.} }

%0 = inproceedings %A = Jeffery, Shawn R. and Franklin, Michael J. and Halevy, Alon Y. %B = Proceedings of the 2008 ACM SIGMOD international conference on Management of data %C = New York, NY, USA %D = 2008 %I = ACM %T = Pay-as-you-go user feedback for dataspace systems %U = http://doi.acm.org/10.1145/1376616.1376701

@inproceedings{conf/sigmod/LeeDDHD04, author = {Dhamankar, Robin and Lee, Yoonkyong and Doan, AnHai and Halevy, Alon Y. and Domingos, Pedro}, title = {iMAP: Discovering Complex Mappings between Database Schemas.}, editor = {Weikum, Gerhard and König, Arnd Christian and Deßloch, Stefan}, booktitle = {SIGMOD Conference}, publisher = {ACM}, year = {2004}, pages = {383-394}, url = {http://www.cs.washington.edu/homes/pedrod/papers/sigmod04.pdf}, isbn = {1-58113-859-8}, keywords = {semantic, srl, mapping, web} }

%0 = inproceedings %A = Dhamankar, Robin and Lee, Yoonkyong and Doan, AnHai and Halevy, Alon Y. and Domingos, Pedro %B = SIGMOD Conference %D = 2004 %I = ACM %T = iMAP: Discovering Complex Mappings between Database Schemas. %U = http://www.cs.washington.edu/homes/pedrod/papers/sigmod04.pdf

@inproceedings{halevymadhavan2003, author = {Halevy, Alon Y. and Madhavan, Jayant}, title = {Corpus-Based Knowledge Representation}, editor = {Gottlob, Georg and Walsh, Toby}, booktitle = {IJCAI-03, Proceedings of the Eighteenth International Joint Conference
on Artificial Intelligence, Acapulco, Mexico, August 9-15, 2003}, publisher = {Morgan Kaufmann}, year = {2003}, pages = {1567-1572}, keywords = {representation, based, corpus, knowledge} }

%0 = inproceedings %A = Halevy, Alon Y. and Madhavan, Jayant %B = IJCAI-03, Proceedings of the Eighteenth International Joint Conference
on Artificial Intelligence, Acapulco, Mexico, August 9-15, 2003 %D = 2003 %I = Morgan Kaufmann %T = Corpus-Based Knowledge Representation

@inproceedings{doan02learning, author = {Doan, AnHai and Madhavan, Jayant and Domingos, Pedro and Halevy, Alon}, title = {Learning to Map between Ontologies on the Semantic Web}, booktitle = {Proceedings to the Eleventh International World Wide}, address = {Honolulu, Hawaii, USA}, year = {2002}, url = {http://www.cs.washington.edu/homes/alon/site/files/glue.pdf}, keywords = {srl, mapping, ontology} }

%0 = inproceedings %A = Doan, AnHai and Madhavan, Jayant and Domingos, Pedro and Halevy, Alon %B = Proceedings to the Eleventh International World Wide %C = Honolulu, Hawaii, USA %D = 2002 %T = Learning to Map between Ontologies on the Semantic Web %U = http://www.cs.washington.edu/homes/alon/site/files/glue.pdf

A data-integration system provides access to a multitude of data sources through a single mediated schema. A key bottleneck in building such systems has been the laborious manual construction of semantic mappings between the source schemas and the mediated schema. We describe LSD, a system that employs and extends current machine-learning techniques to semi-automatically find such mappings. LSD first asks the user to provide the semantic mappings for a small set of data sources, then uses these mappings together with the sources to train a set of learners. Each learner exploits a different type of information either in the source schemas or in their data. Once the learners have been trained, LSD finds semantic mappings for a new data source by applying the learners, then combining their predictions using a meta-learner. To further improve matching accuracy, we extend machine learning techniques so that LSD can incorporate domain constraints as an additional source of knowledge, and develop a novel learner that utilizes the structural information in XML documents. Our approach thus is distinguished in that it incorporates multiple types of knowledge. Importantly, its architecture is extensible to additional learners that may exploit new kinds of information. We describe a set of experiments on several real-world domains, and show that LSD proposes semantic mappings with a high degree of accuracy.

@article{375731, author = {Doan, AnHai and Domingos, Pedro and Halevy, Alon Y.}, title = {Reconciling schemas of disparate data sources: a machine-learning approach}, journal = {SIGMOD Rec.}, publisher = {ACM}, address = {New York, NY, USA}, year = {2001}, volume = {30}, number = {2}, pages = {509--520}, url = {http://portal.acm.org/citation.cfm?id=375731&dl=GUIDE&coll=GUIDE&CFID=75153142&CFTOKEN=89522229}, doi = {http://doi.acm.org/10.1145/376284.375731}, issn = {0163-5808}, keywords = {schema, ol, learning, mapping, data, mining}, abstract = {A data-integration system provides access to a multitude of data sources through a single mediated schema. A key bottleneck in building such systems has been the laborious manual construction of semantic mappings between the source schemas and the mediated schema. We describe LSD, a system that employs and extends current machine-learning techniques to semi-automatically find such mappings. LSD first asks the user to provide the semantic mappings for a small set of data sources, then uses these mappings together with the sources to train a set of learners. Each learner exploits a different type of information either in the source schemas or in their data. Once the learners have been trained, LSD finds semantic mappings for a new data source by applying the learners, then combining their predictions using a meta-learner. To further improve matching accuracy, we extend machine learning techniques so that LSD can incorporate domain constraints as an additional source of knowledge, and develop a novel learner that utilizes the structural information in XML documents. Our approach thus is distinguished in that it incorporates multiple types of knowledge. Importantly, its architecture is extensible to additional learners that may exploit new kinds of information. We describe a set of experiments on several real-world domains, and show that LSD proposes semantic mappings with a high degree of accuracy.} }

%0 = article %A = Doan, AnHai and Domingos, Pedro and Halevy, Alon Y. %C = New York, NY, USA %D = 2001 %I = ACM %T = Reconciling schemas of disparate data sources: a machine-learning approach %U = http://portal.acm.org/citation.cfm?id=375731&dl=GUIDE&coll=GUIDE&CFID=75153142&CFTOKEN=89522229