6533b7d0fe1ef96bd125a17f

RESEARCH PRODUCT

An Approach to Data Quality Evaluation

Janis BicevskisZane BicevskaIvo OditisAnastasija Nikiforova

subject

SQLbusiness.industryComputer sciencemedia_common.quotation_subjectSoftware requirements specificationcomputer.file_formatLinked dataData modelingData integrityData qualityQuality (business)ExecutableSoftware engineeringbusinesscomputermedia_commoncomputer.programming_language

description

This research proposes a new approach to data quality evaluation comprising 3 aspects: (1) data object definition, which quality will be analyzed, (2) quality requirements specification for the data object using Domain Specific Language (DSL), (3) implementation of an executable data quality model that would enable scanning of data object and detect its shortages. Like the Model Driven Architecture (MDA) the data quality modelling is divided into platform independent (PIM) and platform-specific (PSM) models. PIM comprises informal specifications of data quality, PSM describes implementation of data quality model, thus making the data quality model executable. The approbation of the proposed approach provided at least 3 advantages: (1) a graphical data quality model allows definition of data quality by non-IT professionals, (2) the data quality model is not related to the information system that has accumulated data to be analyzed, i.e., this approach allows analysis of an “external” data, and (3) the quality of the data can be described in two abstraction levels - informally using natural language or formally including executable program routines and/or SQL statements.

yearjournalcountryeditionlanguage
2018-10-012018 Fifth International Conference on Social Networks Analysis, Management and Security (SNAMS)
https://doi.org/10.1109/snams.2018.8554915