Challenges for Value-driven Semantic Data Quality Management

Rob Brennan

Knowledge and Data Engineering Group, ADAPT Centre, School of Computer Science and Statistics,

Trinity College Dublin, College Green, Dublin 2, Ireland

Keywords: Data Quality, Semantics, Linked Data, Data Value, Data Management.

Abstract: This paper reflects on six years developing semantic data quality tools and curation systems for both large-

scale social sciences data collection and a major web of data hub. This experience has led the author to

believe in using organisational value as a mechanism for automation of data quality management to deal

with Big Data volumes and variety. However there are many challenges in developing these automated

systems and this discussion paper sets out a set of challenges with respect to the current state of the art and

identifies a number of potential avenues for researchers to tackle these challenges.

1 INTRODUCTION

Data governance aligns organisational goals with the

management of the data assets that data driven

enterprises need to leverage (Brous, 2016). Data

governance thus provides oversight and goal-setting

for data quality management since creating and

maintaining “appropriate” or “fit for use” (i.e. high

quality) data for users is an organisational

imperative (Logan, 2016). 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). In contrast, for

low level data quality metrics, recent advances in

semantic data quality analysis show great promise

(Bertossi, 2013, Feeney, 2016, Kontokostas, 2014)

but methods have not yet emerged to apply them to

traditional databases and semi-structured web data,

where most data growth is centred. Current work on

dataset meta-data standards by the W3C (Maali,

2014) would be a natural basis for business-oriented

quality metadata.

Gartner in 2016 have urged for a more business-led

data governance discipline, highlighting that “it’s all

about the business value”. Moody previously

observed that “100% accurate information is rarely

required in a business context” (Moody, 1999) so it

is impractical (and unprofitable) to blindly try to

achieve “high quality data” across the board (Evan,

2010). Investment in data quality can be seen as

insurance against the risk that your data is not “fit

for use”. In 2013 Tallon spelt out the challenge

“Finding data governance practices that maintain a

balance between value creation and risk exposure is

the new organizational imperative” (Tallon, 2013).

Nonetheless few technologists have taken up this

challenge (Brous, 2016, Yousif, 2015). In part this is

due to a lack of consensus on mathematical models

for the estimation of business value (Viscusi, 2014).

Using value estimates to drive business processes is

common but automated data quality management

toolchains based on value is limited to spot cases

such as quality assessment (Evan, 2005), file

retention management (Wijnhoven, 2014) and data

lifecycle management (Chen, 2005).

This paper presents a survey of recent

developments relevant to developing a new

generation of automated data quality management

systems that are capable of dealing with Big Data

volumes and variety in a way that minimises costs

by using models of the organisational value of data

linked to data quality (section 2). A set of three

research challenges for value-driven data quality

management are then identified along with potential

directions for research that will satisfy these

Brennan, R.

Challenges for Value-driven Semantic Data Quality Management.

DOI: 10.5220/0006387803850392

In Proceedings of the 19th International Conference on Enterprise Information Systems (ICEIS 2017) - Volume 1, pages 385-392

ISBN: 978-989-758-247-9

385

challenges (section 3). Finally in section 4 some

thoughts are presented on the current outlook for

solutions in this space.

2 BACKGROUND EXPERIENCES

AND RELATED WORK

The challenges identified in this paper come from

the author’s experiences developing semantic data

quality tools (Feeney, 2014, Feeney, 2017) and

applying them to large, international data collection

efforts like large, international social science

datasets (Brennan, 2016, Turchin, 2015) or major

linked data hubs (Meehan, 2016). 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). Instead some means

must be developed to focus the attention of

automated tools on the places where they can do the

most good with the least investment of effort. In a

holistic approach that goes beyond the intrinsic or

universal quality measures that the stakeholders

have already rejected, this naturally leads to a desire

for unlocking the greatest value in an organisation’s

data assets. This poses the questions of how can data

value be located and estimated.

Unfortunately, (a) defining what is “appropriate

data” at a business level is both a hard problem and

not adequately addressed by current approaches; and

(b) many dimensions of data value are expressed in

extrinsic data quality measures that depend on

metadata, provenance or usage information that do

not exist within the dataset itself. This leads to the

conjecture that bottom-up data quality tools that

focus on dataset-centric intrinsic quality metrics

such as consistency or integrity will improve over

time but that this progress is only incremental and a

step-change in the effectiveness of data quality

governance requires new methods to driect and

monitor data quality methods and tools based on the

organisational value of data.

Given the explosion in data we are witnessing,

current approaches will not scale to meet the

demand for data that is fit for use. Even with some

progress on tools extrinsic measures like availability

of licensing information (Neumaier, 2016), the real

gains in application of data quality tools will be at

the interface between addressing business needs

(Schork, 2009), supporting domain experts rather

than information architects (Mosley, 2010) and

methods to focus on the available tools and people

on the most relevant data quality issues rather than

wasting effort on uniform metric improvements that

might not even feed into business goals (Evan,

2010). There is a direct parallel between this

situation and the author’s track record on bridging

the gaps in human involvement in semantic mapping

processes (Conroy, 2009, Debruyne, 2013) in

contrast to the majority of the research in ontology

matching which focuses on improving low-level

matching algorithms (Shvaiko, 2013). Another

important influence on the challenges identified for

data quality governance is recent work on semantic

mapping lifecycle governance that uses W3C PROV

as an underlying basis to capture human decision-

making in a machine-readable way (Debruyne,

2015).

Previous work on metadata and business value as

drivers for data quality has focused on pre-semantic

technology for data warehousing (Helfert, 2002,

Shankaranarayan, 2003), organisational decision

support systems (Evan, 2005, Evan, 2010, Schork,

2009, Tallon, 2007, Viscusi, 2014) as opposed to our

challenges for tool automation. There are however

related active research topics like file-retention

strategies based on file metadata (Wijnhoven, 2014)

and autonomic data lifecycle management (Chen,

2005).

A survey of the literature on data governance,

management and lifecycles finds that while data

quality is widely regarded as critical (Brous, 2016,

Khatri, 2010, Tallon, 2013, Weber, 2009), that most

current processes are human rather than machine

oriented (Aiken, 2016, Mosley, 2010). In part this

may be because there are a wide variety of data

lifecycle models but no clear standards (ISO, 2014).

This is influenced by the diversity of data storage

and structuring technologies currently in vogue,

from traditional RDBMS to NoSQL, linked data and

data warehouses to data lakes. Nonetheless the need

for more automated data quality management is

manifest and key to this is how goals are set for

these systems to enable planning, monitoring and

enforcement (Logan, 2016). Underpinning any

automated decision-making will be rich data quality

criteria and the oversight of domain experts such as

ICEIS 2017 - 19th International Conference on Enterprise Information Systems

386

data stewards. These quality criteria must capture

and even predict the links between data assets,

processes, tools, users and data value.

3 THE CHALLENGES

The overall vision of these challenges is to work

towards business-driven, automated semantic data

quality management directed by data value estimates

that becomes more effective over time. Semantics

are at the heart of our approach since they provide

(a) a formal specification model (b) the basis for rich

data quality methods and tools and (c) an effective

data quality integration and interchange technology

when instantiated as enterprise Linked Data.

Our vision first requires the development by the

community of a deeper understanding of how to

define data asset value in a generic and formal way.

Then it will be possible to specify value-driven data

quality criteria, i.e. ways to express the links

between data value, metadata, processes and data

quality methods or tools. Then, this knowledge can

be codified in machine-processable formal models

that support semantic data interchange about data

quality and value in the organisation. This in turn

will enable the development of improved predictive

models and intelligent adaptive data quality systems

for value-driven data quality in digital enterprises. In

a heterogeneous environment the semantic data

interchange models could provide the basis of multi-

vendor interoperability and tool-chain integration.

Each challenge is discussed in a sub-section below,

along with potential approaches for addressing the

challenge.

3.1 Mature Models of Data Value

Although there is a lot of discussion about “data

value”, “information asset valuation”, “data as an

asset” and infonomics at the moment, most of this

discussion is industry-led and does not focus on

formal models of the difficult topic of exactly how

information should be valued. Moody and Walsh

defined seven “laws” of information that explained

its unique behaviour and relation to business value

(Moody and Walsh, 1999) but even that work does

not define the concrete measurement techniques or

metrics. Moody identifies three methods of data

valuation – utility, market price and cost (of

collection) – and concludes that utility is in theory

best but impractical and thus cost-based estimation

is the most effective method. Unfortunately sunk

cost is not a strong candidate for directing future

quality management activities in an agile

environment. Most research on information value

merely seeks to identify dimensions or

characteristics without defining a mathematical

theory of data value.

One guiding research question should be: What

are the fundamental processes and attributes driving

changes in data value over time, and how does this

give rise to patterns of data value development and

diffusion? This would lead to a formal model with

strong explanatory or predictive properties. More

importantly it could act as a baseline for future

research in this under-specified area.

3.1.1 Potential Approaches

There has been no work to date on formal

knowledge models of the data value domain which

limits the application of intelligent systems to data

value management or profiling. The ultimate goal of

such models would be predicting value as well as

assessing it but so far this has proved very context-

dependent (e.g. the value of a specific dataset is a

function of current business goals) and thus hard to

formulate general models. However it is likely that,

as with data quality, there are both extrinsic and

intrinsic measures of value and by calculating the

intrinsic measures it may be possible to estimate the

extrinsic proportion of value

Another area that must be addressed is long-term

validation of models against known data and

business lifecycles. It is possible that usage-based

models of data value, such as already deployed for

file management (Wijnhoven 2014) could be applied

to data assets as a whole. This also corresponds to

the concept of economic value in usage. Such usage-

based models may be based on system logs or

provenance information.

3.2 Linking Data Quality to Data Value

This challenge is about defining value-based data

quality criteria that enable unified quality

governance of datasets and systems. Formal models

of data quality criteria that enable us to better

understand, evaluate and predict the links between

dataset production costs, utility, usage patterns,

quality metrics, metadata, topic domains, workflows,

provenance, and value would enable new value-

driven approaches to data quality governance and

new insights into the location of data value within an

organisation. These machine-processable models

would form the basis for sharing knowledge about

data quality and value throughout the data quality

Challenges for Value-driven Semantic Data Quality Management

387

ecosystem and enable automation of quality

management tasks such as data quality metric

selection, tool selection and orchestration, process or

workflow configuration and quality task

pritorisation, These models would also support new

intelligent data quality applications such as data

asset value profiling or improvement, and decision

support for data quality process design.

Many of the data value dimensions identified in

the literature overlap with data quality dimensions.

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 Zalveri

et al’s recent survey of Linked Data quality

dimensions: accessibility, contextual, dataset

dynamicity, intrinsic, representation and trust

(Zalveri 2015).

3.2.1 Potential Approaches

Linked Data could be used as a unifying technical

foundation for quality criteria specification, dataset

description (via metadata), dataset usage logs and

provenance, and process or tool integration (through

the specification of exchange formats and

lightweight REST-based interfaces).

This parallels the work being carried out at the

W3C on dataset-level metadata and data quality

metric vocabularies and then reused within the

H2020 ALIGNED project to describe the combined

software and data engineering of data-intensive

systems. By creating standardised, reusable semantic

specifications it is possible to build models of a

domain (such as data value) and relate it to

component models describing, for example, data

quality, data lifecycles, business context and

governance roles or processes. Then each sub-model

becomes a basis for data collection and exchange

about a specific dimension of data value, e.g. usage

patterns. The upper or combined model then

becomes the basis for data fusion to determine

overall value.

3.3 Methods to Apply Semantic Data

Quality Tools to Heterogeneous

Data

Effective data driven enterprises require data that is

“fit for use” and must employ active data quality

management of mixed data ecosystems (e.g.

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.

3.3.1 Potential Approaches

Support for dealing with the diversity of real world

data infrastructure could be provided by four

management capabilities: unified dataset metadata

agents, ontology-based data access for quality tools,

and a unified PROV-based log service. Together

these semantic approaches could use the power of

RDF-based data to span multiple local schemata,

provide formal models of semi-structured data

mappings (R2RML-F), reuse rich metadata

specifications and unified data access for multiple

storage technologies. However RDF would only be

required at the data quality management, metadata

and semi-structured data mapping integration points

– existing access to data silos based on end-user

applications or technologies would be unaffected.

This is important to both: (1) be able to deploy these

solutions for real-world data sources and (2) to

ensure that the systems are flexible enough to cope

with diverse data ecosystems rather than being

tailored to an idealised “green-field” deployment of

semantic web technology. This approach follows the

W3C’s Data Activity which envisages complimentary,

connecting pipelines of diverse data formats and

technologies to provide information services.

3.4 Automated Techniques for Value-

Driven Data Quality Management

Satisfying these requirements in the time of the Big

Data deluge requires a shift away from human-

centric processes and requires us to develop new

automated techniques for data quality monitoring,

analysis, and enforcement that assure business value

while minimising human effort. This especially

applies to data quality where the rationale that

human oversight is required for the highest quality

data processes and the limited capabilities of many

traditional data quality tools leads to heavy use of

manual effort in the data quality domain.

3.4.1 Potential Approaches

Automated quality management requires making and

ICEIS 2017 - 19th International Conference on Enterprise Information Systems

388

implementing data quality decisions about data

assets, e.g. selecting quality metrics or processes,

generating quality reports, orchestrating quality

processes or tools (Khatri 2010). This can use

machine reasoning, inference or statistical

approaches based on leveraging knowledge models

of: the data quality domain (Radulovic, 2016),

threats to data quality and catalogues of best practice

(Foley, 2011), how data value can be expressed in

dataset metadata (Helfert, 2002, Wijnhoven, 2014)

and extrinsic data quality metrics (Viscusi, 2014).

Hence the presence of formal models linking data

quality and value enable automated decision-making

or decision support for data owners such as Data

Stewards.

Specific technology performance curves could be

developed for data quality processes and tools that,

in conjunction with the knowledge models, support

data quality planning and prediction. Success will be

measured by statistical analysis of the model’s

predictions using historical studies of data quality

(hindcasting).

By building on Chen’s work on value-based

autonomic data lifecycle management (Chen, 2005)

and Even et al.’s approach to value-based data

quality management for organisational decision

support (Evan, 2010), it may be possible to support

automated selection, prioritisation and orchestration

of data quality tools.

3.5 Standardised Data Quality

Management Architecture

A common architecture and standardised integration

reference points will make the latest advances in

semantic data quality tools available to traditional

databases and semi-structured web data via

ontology-based data access (challenge 3), common

metadata standards and semantic mappings. The

reference architecture should enable a semi-

supervised data quality control loop (challenge 4).

The behaviour of the control loop goals would be set

by data quality criteria expressed in terms of a

governance model that links value and data quality

(challenge 2). This will address the gap in the state

in the art whereby most research looks at individual

quality tools outside of their deployment context and

without reference to any business context.

3.5.1 Potential Approaches

These integration reference points could bring

together disparate semantic quality reporting,

metadata and data quality vocabularies into a

coherent whole, enabling multi-vendor solutions.

This would extend the deployment scope of

semantic data quality tools by defining a new

approach to quality-centric ontology-based data

access (OBDA), where to date only consistency

measures have been addressed [Console14]. New

mechanisms for data quality management of

semantic mappings and semi-structured data could

be developed to allow semantic quality approaches

to be applied to semi-structured data.

An exemplar architecture for automated, value-

driven semantic data quality management is

sketched in figure 1 below. Support for dealing with

the diversity of real world data infrastructure will be

provided by four management capabilities: unified

dataset metadata agents, unified ontology-based data

access for quality tools, and a unified PROV-based

log service. Together these semantic approaches use

the power of RDF-based data to span multiple local

schemata, provide formal models of semi-structured

data mappings (R2RML-F), reuse rich metadata

specifications and unified data access for multiple

storage technologies. However RDF is only required

at the data quality management, metadata and semi-

structured data mapping integration reference points

– existing access to data silos based on end-user

applications or technologies are unaffected. This is

important to both: (1) be able to deploy solutions for

real-world data sources and (2) to ensure that the

integration reference point and data quality criteria

specifications are flexible enough to cope with

diverse data ecosystems rather than being tailored to

an idealised “green-field” deployment of semantic

web technology.

The architecture components shown in figure 1 are

as follows:

Automated Data Quality System: this will

monitor, analyse and enforce data quality within the

data quality management system based on the value-

driven autonomic data lifecycle approach of Chen

[Chen05].

Semantic Data Quality Tools from the state of

the art such as TCD’s Dacura Quality Service for

OWL-based validation (Feeney, 2017) and AKSW’s

RDFUnit tool for SPARQL and SHACL-based data

unit testing (Kontokostas, 2014).

Data Access for Quality Management: Access to

RDBMS for metadata agents and semantic data

quality tools would be based on the mature and

highly performant ontology-based data access

platforms such as OnTop (Calvanese et al, 2016).

Dataset Log Agents will convert, create and

maintain dataset usage and governance information

using the W3C’s PROV standard.

Challenges for Value-driven Semantic Data Quality Management

389

Figure 1: Exemplar Automated Value-Driven Semantic Data Management Architecture.

Dataset Metadata Agents could populate and

maintain DataValue extensions of emerging

metadata standards like DataID and capture

metadata fields relevant to calculating data value

such as key entities and dataset provenance.

4 SUMMARY AND OUTLOOK

This paper discussed significant challenges facing

data quality researchers and the big data industry as

it aims to tackle the dual goals of controlling data

quality costs and engineering those systems to be

flexible enough to support agile management

decision making by data service providers and their

customers in the face of the increased scalability

demands of Big Data systems.

Most of the development of value-driven systems

is currently outside of computer science or

informatics academic research and is led by industry

specialists such as Doug Laney of Gartner. There is

a parallel thread of academic research on knowledge

management and organisational impact that emerges

from the business schools or economics

departments. However these three threads must

come together if we are to engineer value-driven

systems. This requires bridging the gap between

human understanding of business needs and low-

level data lifecycle tools. Hence semantics or formal

knowledge models are ideally placed to play a

significant role in future systems. This compliments

the W3C’s standardisation role on knowledge-based,

data-centric systems.

ACKNOWLEDGEMENTS

This research has received funding from the ADAPT

Centre for Digital Content Technology, funded

under the SFI Research Centres Programme (Grant

13/RC/2106) and co-funded by the European

Regional Development Fund.

The author also wants to thank the reviewers for

their many suggestions for improving the final

version of this paper.

REFERENCES

Ahituv, N., 1989, Assessing the value of information:

problems and approaches. In: DeGross,

J.I.,Henderson, J.C., Konsynski, B.R. (eds.)

International Conference on Information Systems

(ICIS 1989). pp. 315–325. Boston, Massachusetts.

ICEIS 2017 - 19th International Conference on Enterprise Information Systems

390

Aiken, P., 2016, EXPERIENCE: Succeeding at Data

Management—BigCo Attempts to Leverage Data,

Journal of Data and Information Quality (JDIQ),

Volume 7 Issue 1.

Bertossi, L. and Bravo, L., 2013, Generic and Declarative

Approaches to Data Cleaning: Some Recent

Developments, Handbook of Data Quality: Research

and Practice, Shazia Sadiq (Ed), Spriner, ISBN 978-3-

642-36256-9, 2013.

Brennan, R. , Feeney, K. Mendel-Gleason, G. Bozic, B.

Turchin, P. Whitehouse, H. Francois, P. Currie, T. E.

Grohmann, S., 2011, Building the Seshat Ontology for

a Global History Databank, LNCS , Extended

Semantic Web Conference, Heraklion, 29th May - 2nd

June, edited by Harald Sack - Eva Blomqvist -

Mathieu d'Aquin - Chiara Ghidini - Simone Paolo

Ponzetto - Christoph Lange , (9678), Springer, 2016,

pp693 – 708.

Calvanese, D., Cogrel, B. , Komla-Ebri, S. Kontchakov, R.

Lanti, D. Rezk, M. Rodriguez-Muro, M. Xiao, G.,

2016, Ontop: Answering SPARQL Queries over

Relational Databases, (Accepted), Semantic Web

Journal, Available at: http://www.semantic-web-

journal.net/content/ontop-answering-sparql-queries-

over-relational-databases-1 [Accessed 09 March 2017]

Chen, Y., 2005, Information valuation for Information

Lifecycle Management, Proc IEEE International

Conference on Autonomic Computing pp: 135-146,

DOI 10.1109/ICAC.2005.35.

Console, M. and Lenzerini, M., 2014, Data Quality in

Ontology-Based Data Access:The Case of

Consistency, Proceedings of the Twenty-Eighth AAAI

Conference on Artificial Intelligence.

Debruyne, C., Walshe, B., O'Sullivan, D., 2015, Towards

a Project Centric Metadata Model and Lifecycle for

Ontology Mapping Governance, 17th International

Conference on Information Integration and Web-based

Applications & Services, Brussels, Belgium, 11-13

December , edited by Maria Indrawan-Santiago,

Matthias Steinbauer, A Min Tjoa, Ismail Khalil,

Gabriele Anderst-Kotsis , ACM, pp356 – 365.

Vander Sande, M., Colpaert, P., Verborgh, R., Mannens,

E., and Van De Walle, R., 2014 , RML : A Generic

Language for Integrated RDF Mappings of

Heterogeneous Data.

Even, A., and Shankaranarayanan, G., 2005, Value-Driven

Data Quality Assessment, in Proceedings of the 10th

International Conference on Information Quality,

Cambridge, MA, USA.

Evan, A., 2010, Evaluating a model for cost-effective data

quality management in a real-world CRM setting,

Decision Support Systems 50(1):152-163, DOI:

10.1016/j.dss.2010.07.011.

Feeney, K., O'Sullivan, D., Tai, W. and Brennan, R. 2014,

Improving curated web-data quality with structured

harvesting and assessment, International Journal on

Semantic Web and Information Systems 10(2).

Feeney, K. Mendel-Gleason G. and Brennan, R., 2017,

Linked data schemata: fixing unsound foundations,

Semantic Web Journal, Accepted, Available at:

http://www.semantic-web-journal.net/content/linked-

data-schemata-fixing-unsound-foundations-0

[Accessed 09 March 2017]

Foley, S., and Fitzgerald, W., 2011, Management of

security policy configuration using a Semantic Threat

Graph approach, Journal of Computer Security, vol.

19, no. 3, pp. 567-605.

Helfert, M., and Herrmann, C., 2002, Proactive Data

Quality Management for Data Warehouse Systems - A

Metadata based Data Quality System. In: 4th

International Workshop on Design and Management

of Data Warehouses (DMDW), Toronto, Canada.

ISO/IEC JTC 1, 2014, Information Technology, Big Data,

Preliminary Report 2014. Available at:

http://www.iso.org/iso/big_data_report-jtc1.pdf

[Accessed 09 March 2017]

Janssen, M. and Vilminko-Heikkinen, R., 2016,

Coordinating Decision-Making in Data Management

Activities: A Systematic Review of Data Governance

Principles, EGOV 2016, LNCS 9820, pp. 115–

125.DOI: 10.1007/978-3-319-44421-5_9.

Khatri, V., Brown, C.V., 2010, Designing data

governance. Communications of the ACM 53(1), 148–

152.

Kontokostas, D. , Westphal, P. , Auer, S., Hellmann, S.,

Lehmann, J. , Cornelissen, r., Zaveri, a., 2014, Test-

driven Evaluation of Linked Data Quality,

Proceedings of the 23rd International Conference on

World Wide Web, pp747-758.

Logan, D. 2016, Ten Steps to Information Governance,

Gartner Report G00296492. Available at:

https://www.edq.com/resources/analyst-

reports/gartner-10-steps-to-data-governance/

[Accessed 09 March 2017]

Maali, F., Erickson, J., (Eds.), 2014, Data Catalog

Vocabulary (DCAT), W3C Recommendation,

Available at https://www.w3.org/TR/vocab-dcat/

Meehan, A. Kontokostas, D. Freudenberg, M. Brennan, R.

O'Sullivan, D., 2016, Validating Interlinks between

Linked Data Datasets with the SUMMR Methodology,

ODBASE 2016 - The 15th International Conference

on Ontologies, DataBases, and Applications of

Semantics, Rhodes, Greece, , Springer Verlag.

Mendel-Gleason, G. Feeney K. and Brennan, R., 2015,

Ontology Consistency and Instance Checking for Real

World Linked Data, 2nd Workshop on Linked Data

Quality, Slovenia.

Moody, D. and Walsh, P., 1999, Measuring The Value Of

Information: An Asset Valuation Approach, Proc.

Seventh European Conference on Information Systems

(ECIS’99).

Mosley, M., Brackett, M.,, Earley, S., and Henderson, D.,

(Eds), 2010, The DMA Guide to the Data

Management Body of Knowledge (1st Ed.), Technics

Publications LLC, 2010, ISBN 978-9355040-2-3.

Neumaier, S., Umbrich, J., Polleres, A., 2016, Automated

Quality Assessment of Metadata across Open Data

Portals, Journal of Data and Information Quality

(JDIQ), Vol. 8 Issue 1, DOI:10.1145/2964909.

Challenges for Value-driven Semantic Data Quality Management

391

Schork, R., 2009, Integrating Business and Technical

Metadata to Support Better Data Quality, MIT

Information Quality Industry Symposium. Available at

http://mitiq.mit.edu/IQIS/Documents/CDOIQS_20097

7/Papers/02_02_1C-2.pdf [Accessed 09 March 2017]

Solanki, M., Bozic, B., Freudenberg, M., Kontokostas, D.,

Dirschl C., and Brennan,R., 2016, Enabling combined

Software and Data engineering at Web-scale: The

ALIGNED suite of Ontologies, ISWC 2016: The

Semantic Web – ISWC 2016 pp 195-203.

Radulovic, F., Mihindukulasooriya, N., García-Castro, R.

and Gómez-Pérez, A., 2016, A comprehensive quality

model for Linked Data, Accepted, Semantic Web

Journal, Available at: http://www.semantic-web-

journal.net/system/files/swj1488.pdf [Accessed 09

March 2017]

Shankaranarayan, G., Ziad, M., and Wang, R. Y., 2003,

Managing Data Quality in Dynamic Decision

Environments: An Information Product Approach,

Journal of Database Management (JDM), 14 (4).

Shvaiko, P., Euzenat, J., 2013, Ontology Matching: State

of the Art and Future Challenges, IEEE Trans.

Knowledge and Data Engineering 25(1).

Tallon, P. and Scannell, P., 2007, Information Lifecycle

Management, Communications of the ACM, Volume

50 Issue 11, pp. 65-69.

Tallon, P., 2013, Corporate Governance of Big Data:

Perspectives on Value, Risk, and Cost, IEEE

Computer, Vol 46, Issue: 6.

Turchin, P., Brennan, R., Currie, T., Feeney, K., Francois,

P., Hoyer, D., Manning, J., Marciniak, A., Mullins, D.,

Palmisano, A., Peregrine, P., Turner, E.A.L.,

Whitehouse, H., 2015, Seshat: The Global History

Databank, Cliodynamics: The Journal of Quantitative

History and Cultural Evolution, 6, (1), 2015, p77 –

107.

Viscusi, G., and Batini, C., 2014, Digital Information

Asset Evaluation: Characteristics and Dimensions", in

Smart Organizations and Smart Artifacts: Fostering

Interaction Between People, Technologies and

Processes, pp. 77-86, Springer International

Publishing. DOI:10.1007/978-3-319-07040-7_9.

Weber, K., Otto, B. and Osterle, H., 2009, One size does

not fit all—a contingency approach to data

governance. J.Data Inf. Qual. 1(1), 1–27.

Wijnhoven, F., Amrit, C. and Dietz, P., 2014, Value-based

File Retention: File Attributes as File Value and

Information Waste Indicators, ACM journal of data

and information quality, 4 (4). 15 -. ISSN 1936-1955.

Yousif, M., 2015, The Rise of Data Capital, IEEE Cloud

Computing.

Zaveri, A., Rula, A., Maurino, A., Pietrobon, R.,

Lehmann, J., and Auer, S., 2015,Quality Assessment

for Linked Data: A Survey, Semantic Web –

Interoperability, Usability, Applicability, IOS Press,

ISSN: 1570-0844.

ICEIS 2017 - 19th International Conference on Enterprise Information Systems

392