User Feedback as a First Class Citizen in Information Integration Systems

User feedback is gaining momentum as a means of addressing the diculties underlying information integration tasks. It can be used to assist users in building information integration systems and to improve the quality of existing systems, e.g., in dataspaces. Existing proposals in the area are conned to specic integration sub-problems considering a specic kind of feedback sought, in most cases, from a single user. We argue in this paper that, in order to maximize the benets that can be drawn from user feedback, it should be considered and managed as a rst class citizen. Accordingly, we present generic operations that underpin the management of feedback within information integration systems, and that are applicable to feedback of dierent kinds, potentially supplied by multiple users with dierent expectations. We present preliminary solutions that can be adopted for realizing such operations, and sketch a research agenda for the information integration community.

[1]  Pavel Berkhin,et al.  A Survey of Clustering Data Mining Techniques , 2006, Grouping Multidimensional Data.

[2]  K. Selçuk Candan,et al.  Feedback-driven result ranking and query refinement for exploring semi-structured data collections , 2010, EDBT '10.

[3]  Anil K. Jain,et al.  Data clustering: a review , 1999, CSUR.

[4]  Alon Y. Halevy,et al.  Pay-as-you-go user feedback for dataspace systems , 2008, SIGMOD Conference.

[5]  Aristides Gionis,et al.  Clustering aggregation , 2005, 21st International Conference on Data Engineering (ICDE'05).

[6]  AnHai Doan,et al.  Matching Schemas in Online Communities: A Web 2.0 Approach , 2008, 2008 IEEE 24th International Conference on Data Engineering.

[7]  David Maier,et al.  From databases to dataspaces: a new abstraction for information management , 2005, SGMD.

[8]  Joann J. Ordille,et al.  Data integration: the teenage years , 2006, VLDB.

[9]  Marek J. Sergot,et al.  A logic-based calculus of events , 1989, New Generation Computing.

[10]  Norman W. Paton,et al.  Feedback-based annotation, selection and refinement of schema mappings for dataspaces , 2010, EDBT '10.

[11]  Renée J. Miller,et al.  Muse: Mapping Understanding and deSign by Example , 2008, 2008 IEEE 24th International Conference on Data Engineering.

[12]  Abhinandan Das,et al.  Google news personalization: scalable online collaborative filtering , 2007, WWW '07.

[13]  AnHai Doan,et al.  Mass Collaboration Systems on the World-Wide Web , 2010 .

[14]  Partha Pratim Talukdar,et al.  Automatically incorporating new sources in keyword search-based data integration , 2010, SIGMOD Conference.

[15]  Jeffrey F. Naughton,et al.  Efficiently incorporating user feedback into information extraction and integration programs , 2009, SIGMOD Conference.

[16]  Koby Crammer,et al.  Learning to create data-integrating queries , 2008, Proc. VLDB Endow..

[17]  Alon Y. Halevy,et al.  Data integration with uncertainty , 2007, The VLDB Journal.

[18]  Taghi M. Khoshgoftaar,et al.  A Survey of Collaborative Filtering Techniques , 2009, Adv. Artif. Intell..