Guildlines of Data Quality Issues for Data Integration in the
Context of the TPC-DI Benchmark

D 2017

Guildlines of Data Quality Issues for Data Integration in the Context of the TPC-DI Benchmark

YANG, Qishan, Mouzhi GE and Markus HELFERT

Basic information

Original name

Guildlines of Data Quality Issues for Data Integration in the Context of the TPC-DI Benchmark

Authors

YANG, Qishan (156 China), Mouzhi GE (156 China, guarantor, belonging to the institution) and Markus HELFERT (276 Germany)

Edition

Porto, Portugal, Proceedings of the 19th International Conference on Enterprise Information Systems, p. 135-144, 10 pp. 2017

Publisher

SciTePress

Other information

Language

English

Type of outcome

Stať ve sborníku

Field of Study

10201 Computer sciences, information science, bioinformatics

Confidentiality degree

není předmětem státního či obchodního tajemství

Publication form

printed version "print"

References:

Springer, indexed by SCOPUS, WoS, DBLP

RIV identification code

RIV/00216224:14330/17:00096406

Organization unit

Faculty of Informatics

ISBN

978-989-758-247-9

DOI

http://dx.doi.org/10.5220/0006334301350144

UT WoS

000697605900013

Keywords in English

Data integration; Data quality; ETL; TPC-DI benchmark

Abstract

V originále

Nowadays, many business intelligence or master data management initiatives are based on regular data integration, since data integration intends to extract and combine a variety of data sources, it is thus considered as a prerequisite for data analytics and management. More recently, TPC-DI is proposed as an industry benchmark for data integration. It is designed to benchmark the data integration and serve as a standardisation to evaluate the ETL performance. There are a variety of data quality problems such as multi-meaning attributes and inconsistent data schemas in source data, which will not only cause problems for the data integration process but also affect further data mining or data analytics. This paper has summarised typical data quality problems in the data integration and adapted the traditional data quality dimensions to classify those data quality problems. We found that data completeness, timeliness and consistency are critical for data quality management in data integration, and data consistency should be further defined in the pragmatic level. In order to prevent typical data quality problems and proactively manage data quality in ETL, we proposed a set of practical guidelines for researchers and practitioners to conduct data quality management in data integration.

Citovat

YANG, Qishan, Mouzhi GE and Markus HELFERT. Guildlines of Data Quality Issues for Data Integration in the Context of the TPC-DI Benchmark. In Proceedings of the 19th International Conference on Enterprise Information Systems. Porto, Portugal: SciTePress, 2017, p. 135-144. ISBN 978-989-758-247-9. Available from: https://dx.doi.org/10.5220/0006334301350144.

@inproceedings{1377092,
   author = {Yang, Qishan and Ge, Mouzhi and Helfert, Markus},
   address = {Porto, Portugal},
   booktitle = {Proceedings of the 19th International Conference on Enterprise Information Systems},
   doi = {http://dx.doi.org/10.5220/0006334301350144},
   keywords = {Data integration; Data quality; ETL; TPC-DI benchmark},
   howpublished = {tištěná verze "print"},
   language = {eng},
   location = {Porto, Portugal},
   isbn = {978-989-758-247-9},
   pages = {135-144},
   publisher = {SciTePress},
   title = {Guildlines of Data Quality Issues for Data Integration in the Context of the TPC-DI Benchmark},
   url = {http://iceis.org/},
   year = {2017}
}

TY  - JOUR
ID  - 1377092
AU  - Yang, Qishan - Ge, Mouzhi - Helfert, Markus
PY  - 2017
TI  - Guildlines of Data Quality Issues for Data Integration in the Context of the TPC-DI Benchmark
PB  - SciTePress
CY  - Porto, Portugal
SN  - 9789897582479
KW  - Data integration
KW  - Data quality
KW  - ETL
KW  - TPC-DI benchmark
UR  - http://iceis.org/
N2  - Nowadays, many business intelligence or master data management initiatives are based on regular data integration, since data integration intends to extract and combine a variety of data sources, it is thus considered as a prerequisite for data analytics and management. More recently, TPC-DI is proposed as an industry benchmark for data integration. It is designed to benchmark the data integration and serve as a standardisation to evaluate the ETL performance. There are a variety of data quality problems such as multi-meaning attributes and inconsistent data schemas in source data, which will not only cause problems for the data integration process but also affect further data mining or data analytics. This paper has summarised typical data quality problems in the data integration and adapted the traditional data quality dimensions to classify those data quality problems. We found that data completeness, timeliness and consistency are critical for data quality management in data integration, and data consistency should be further defined in the pragmatic level. In order to prevent typical data quality problems and proactively manage data quality in ETL, we proposed a set of practical guidelines for researchers and practitioners to conduct data quality management in data integration.
ER  -

YANG, Qishan, Mouzhi GE and Markus HELFERT. Guildlines of Data Quality Issues for Data Integration in the Context of the TPC-DI Benchmark. In \textit{Proceedings of the 19th International Conference on Enterprise Information Systems}. Porto, Portugal: SciTePress, 2017, p.~135-144. ISBN~978-989-758-247-9. Available from: https://dx.doi.org/10.5220/0006334301350144.

Detailed Information on Publication Record