Quality assessment for Linked Data: A Survey

Abstract

The development and standardization of semantic web technologies has resulted in an unprecedented volume of data being published on the Web as Linked Data (LD). However, we observe widely varying data quality ranging from extensively curated datasets to crowdsourced and extracted data of relatively low quality. In this article, we present the results of a systematic review of approaches for assessing the quality of LD. We gather existing approaches and analyze them qualitatively. In particular, we unify and formalize commonly used terminologies across papers related to data quality and provide a comprehensive list of 18 quality dimensions and 69 metrics. Additionally, we qualitatively analyze the 30 core approaches and 12 tools using a set of attributes. The aim of this article is to provide researchers and data curators a comprehensive understanding of existing work, thereby encouraging further experimentation and development of new approaches focused towards data quality, specifically for LD.

Challenges for Value-driven Semantic Data Quality Management

“…For example Ahituv (Ahituv, 1989) suggests: timeliness (dimensions: recency, response time, and frequency), contents (dimensions: accuracy, relevance, level of aggregation and exhaustiveness), format (dimensions: media, color, structure, presentation), and cost. Compare these with Zaveri et al's recent survey of Linked Data quality dimensions: accessibility, contextual, dataset dynamicity, intrinsic, representation and trust (Zaveri 2015).…”

“…relational databases, linked data and semi-structured data like csv, json and xml) while linking quality actions to business value. Despite recent progress and the emergence of both commercial and research tools (Zaveri, 2015), semantic data quality tool researchers must address the fact that the majority of the world's data is not stored in RDF graphs. This limits the applicability and impact of their tools and methods.…”

“…Interactions with dataset stakeholders over a number of years have suggested that despite the advantages of semantic data quality approaches, e.g. more expressive schema (Mendel-Gleason, 2015), that the intrinsic quality metrics that the majority of such tools focus on (Zaveri, 2015) are not the locus of business or organisational value in the datasets (Evan, 2005). Even the best data quality processes and tools require human oversight to be most effective (Brous, 2016) and as the number and variety of datasets increases (especially in a world dominated by Big Data) it is not scaleable to try and improve data quality uniformly (Evan, 2010).…”

“…To date, bridging the gap between business priorities and data quality management has been hampered by factors such as: the focus of data quality tools on low-level intrinsic quality measures (e.g. syntactic validity) (Zaveri, 2015), the multi-faceted nature of data quality (Radulovic, 2016), an increasingly diverse data technology ecosystem built on siloed tool-chains (Khatri, 2010), and the lack of business-oriented metadata (Schork, 2009). Automated support for data stakeholder or steward governance of data quality is immature due to the lack of standardisation of integration points for big data quality control systems (ISO, 2014).…”

Enriching and improving the quality of linked data with GIS

“…Moreover, from the practical point of view this assumption is often essential [14]. Other metrics of linked data quality include coherence of links to other resources or consistency with regard to implicit information [21]. Recent studies pointed out problems with linked data quality concerning non-standardized and variable representation, inconsistency, and lack of interoperability [18,25].…”