Qualitative Effects of Knowledge Rules and User Feedback in Probabilistic Data Integration

Maurice van Keulen, Ander de Keijzer

Research output: Contribution to journalArticleAcademicpeer-review

38 Downloads (Pure)

Abstract

In data integration efforts, portal development in particular, much development time is devoted to entity resolution. Often advanced similarity measurement techniques are used to remove semantic duplicates or solve other semantic conflicts. It proves impossible, however, to automatically get rid of all semantic problems. An often-used rule of thumb states that about 90% of the development effort is devoted to semi-automatically resolving the remaining 10% hard cases. In an attempt to significantly decrease human effort at data integration time, we have proposed an approach that strives for a ‘good enough’ initial integration which stores any remaining semantic uncertainty and conflicts in a probabilistic database. The remaining cases are to be resolved with user feedback during query time. The main contribution of this paper is an experimental investigation of the effects and sensitivity of rule definition, threshold tuning, and user feedback on the integration quality. We claim that our approach indeed reduces development effort — and not merely shifts the effort—by showing that setting rough safe thresholds and defining only a few rules suffices to produce a ‘good enough’ initial integration that can be meaningfully used, and that user feedback is effective in gradually improving the integration quality.
Original languageUndefined
Article number10.1007/s00778-009-0156-z
Pages (from-to)1191-1217
Number of pages27
JournalVLDB journal
Volume18
Issue number5
DOIs
Publication statusPublished - Oct 2009

Keywords

  • DB-SDI: SCHEMA AND DATA INTEGRATION
  • EWI-15672
  • Data integration - Entity resolution - Uncertain databases - Data quality - User feedback
  • IR-67796
  • METIS-265217

Cite this

@article{f28b9ea7580c4e258a350c0754f91f49,
title = "Qualitative Effects of Knowledge Rules and User Feedback in Probabilistic Data Integration",
abstract = "In data integration efforts, portal development in particular, much development time is devoted to entity resolution. Often advanced similarity measurement techniques are used to remove semantic duplicates or solve other semantic conflicts. It proves impossible, however, to automatically get rid of all semantic problems. An often-used rule of thumb states that about 90{\%} of the development effort is devoted to semi-automatically resolving the remaining 10{\%} hard cases. In an attempt to significantly decrease human effort at data integration time, we have proposed an approach that strives for a ‘good enough’ initial integration which stores any remaining semantic uncertainty and conflicts in a probabilistic database. The remaining cases are to be resolved with user feedback during query time. The main contribution of this paper is an experimental investigation of the effects and sensitivity of rule definition, threshold tuning, and user feedback on the integration quality. We claim that our approach indeed reduces development effort — and not merely shifts the effort—by showing that setting rough safe thresholds and defining only a few rules suffices to produce a ‘good enough’ initial integration that can be meaningfully used, and that user feedback is effective in gradually improving the integration quality.",
keywords = "DB-SDI: SCHEMA AND DATA INTEGRATION, EWI-15672, Data integration - Entity resolution - Uncertain databases - Data quality - User feedback, IR-67796, METIS-265217",
author = "{van Keulen}, Maurice and {de Keijzer}, Ander",
note = "eemcs-eprint-15672",
year = "2009",
month = "10",
doi = "10.1007/s00778-009-0156-z",
language = "Undefined",
volume = "18",
pages = "1191--1217",
journal = "VLDB journal",
issn = "1066-8888",
publisher = "Springer",
number = "5",

}

Qualitative Effects of Knowledge Rules and User Feedback in Probabilistic Data Integration. / van Keulen, Maurice; de Keijzer, Ander.

In: VLDB journal, Vol. 18, No. 5, 10.1007/s00778-009-0156-z, 10.2009, p. 1191-1217.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Qualitative Effects of Knowledge Rules and User Feedback in Probabilistic Data Integration

AU - van Keulen, Maurice

AU - de Keijzer, Ander

N1 - eemcs-eprint-15672

PY - 2009/10

Y1 - 2009/10

N2 - In data integration efforts, portal development in particular, much development time is devoted to entity resolution. Often advanced similarity measurement techniques are used to remove semantic duplicates or solve other semantic conflicts. It proves impossible, however, to automatically get rid of all semantic problems. An often-used rule of thumb states that about 90% of the development effort is devoted to semi-automatically resolving the remaining 10% hard cases. In an attempt to significantly decrease human effort at data integration time, we have proposed an approach that strives for a ‘good enough’ initial integration which stores any remaining semantic uncertainty and conflicts in a probabilistic database. The remaining cases are to be resolved with user feedback during query time. The main contribution of this paper is an experimental investigation of the effects and sensitivity of rule definition, threshold tuning, and user feedback on the integration quality. We claim that our approach indeed reduces development effort — and not merely shifts the effort—by showing that setting rough safe thresholds and defining only a few rules suffices to produce a ‘good enough’ initial integration that can be meaningfully used, and that user feedback is effective in gradually improving the integration quality.

AB - In data integration efforts, portal development in particular, much development time is devoted to entity resolution. Often advanced similarity measurement techniques are used to remove semantic duplicates or solve other semantic conflicts. It proves impossible, however, to automatically get rid of all semantic problems. An often-used rule of thumb states that about 90% of the development effort is devoted to semi-automatically resolving the remaining 10% hard cases. In an attempt to significantly decrease human effort at data integration time, we have proposed an approach that strives for a ‘good enough’ initial integration which stores any remaining semantic uncertainty and conflicts in a probabilistic database. The remaining cases are to be resolved with user feedback during query time. The main contribution of this paper is an experimental investigation of the effects and sensitivity of rule definition, threshold tuning, and user feedback on the integration quality. We claim that our approach indeed reduces development effort — and not merely shifts the effort—by showing that setting rough safe thresholds and defining only a few rules suffices to produce a ‘good enough’ initial integration that can be meaningfully used, and that user feedback is effective in gradually improving the integration quality.

KW - DB-SDI: SCHEMA AND DATA INTEGRATION

KW - EWI-15672

KW - Data integration - Entity resolution - Uncertain databases - Data quality - User feedback

KW - IR-67796

KW - METIS-265217

U2 - 10.1007/s00778-009-0156-z

DO - 10.1007/s00778-009-0156-z

M3 - Article

VL - 18

SP - 1191

EP - 1217

JO - VLDB journal

JF - VLDB journal

SN - 1066-8888

IS - 5

M1 - 10.1007/s00778-009-0156-z

ER -