The Potential Impact Of Linked Data

Published Date: 02 Nov 2017

This Document discusses the evolving semantic web and the theoretical process of linked data, with particular focus on the concept, the potential uses in the future for linked data, and discussion of current projects that are involved with the linked open data initiative including the growth it has seen over recent years. The report also discusses security, and the potential legal and ethical concerns. Additional comment is given to an independent study that was open to the public which aimed at gauging public awareness of the potential risks, benefits, and impact of linked data. The survey identifies the employment fields of the participants and their involvement with the structure of data within their organisation, and assesses whether or not they understand the potential benefits and risks of open data.

Methodology 600w

The research, on which this paper is based combines both a quantitative and qualitative assessment of the impact of linked data. The document will take into account, knowledge from several sources, including secondary research articles produced as part of other academic studies into the impact of linked data, and an independent study aimed at gauging public awareness of the concept.

Hannon (2008) suggests that, using this two tiered methodology, and a combination of the positivistic and phenomenological approaches, it will be possible to gauge the current public awareness by the use of a survey, and using case study observation it will be possible to assess the impact of linked data in the real world.

Collis & Hussey (2003) suggest that, research methodologies are usually divided into two main types these are qualitative and quantitative. This document is based on a combination of the two approaches and gathers information from both a primary and secondary data collection.

The Secondary Research Element is based on findings from current academic papers on the concept of linked and open data, this will include, discussion on the potential uses of linked and open data, existing projects, including current participation levels in the open data initiative (ODI). Further discussion is enforced by academic articles centred on the security/legal and ethical concerns, the data standards required, and how businesses can profit from linked data through the use of evolving theoretical profit models.

The Primary Research Element is based on the results from a survey aimed at gauging public awareness of the potential risks and benefits of linked data. The survey includes a series of multiple choice questions aimed at checking a number of different factors, including, sex, age, employed position, employment area, and how much responsibility a particular employee has over the data within their organisation, and whether or not they understand the benefits and risks of linked data. The data gathered from this study will provide an indicator as to whether there is a gap in public understanding.

To make the study achievable, a certain amount of thought had to be given to the information that could be asked for, and whether the results that would be returned would yield anything tangible. See Appendix A for the final survey list of questions used in the study.

Survey Scope:

The questions asked in the survey, were designed in such a way that the obtained results would offer quantifiable evidence of just how many members of the public know about the impact of linked data and how it is affecting them directly "right now". But equally important, is the results obtained from the study, help identify the common themes of where there is a more significant lack of awareness. Because the survey was equally open to business as it was the general public, provision was given to being able to account for participant employment area, and position within the company, the survey prompts participants for how much as a percentage value of their time spent at work, if any, is concerned with the administration, interpretation and structuring of data. And if so are they aware of the potential implications and benefits of linked and open data.

Literature Review: 2200w

What is the Semantic Web? 500w

Concept:

A report produced by the IEEE, by Philippe Cudré-Mauroux (2012) of MIT, explains the history of the semantic web ideology, and states that during the early 1960’s, the concept of a "Semantic Network" was initially introduced by scientist Allan M. Collins, linguist M. Ross Quillian, & psychologist Elizabeth F. Loftus, as a "knowledge representation model". It wasn’t until 1998, that the term Semantic Web was acknowledged by the inventor of the World Wide Web, Tim Berners - Lee. It was and still is Berners – Lee’s vision to enhance the existing Web infrastructure to accommodate raw data. To achieve this, Berners - Lee suggested a huge linked global data network that could be understandable and consumable by machines. Unlike the existing web which is hypertext based, and indicates links to other texts by hyperlink, the Semantic Web, is based upon a "hyper-data" approach, which indicates data objects pointing to other relevant data objects across the Web using specific ontology’s and semantics. (Maroux 2012)

Herman (2009) suggests that, the adoption of the fore mentioned approach to adapt the web infrastructure has yielded many changes over recent years in the way that we now consume and use data. Following specific ontology’s (which are detailed later in this document), it is now possible, using the RDF framework (also detailed later), to publish data that is open, and un-interpreted (Raw) which can be searched for and consumed by others using a myriad of interface methods.

The potential benefits of linked and open data:

The adaptation of the web infrastructure to accommodate the "Hyper-Data" approach has opened up many potential benefits to society. This is identified in a publication by F. Bauer (2006) the paper highlights the benefits of linked data. In his report he states that, a real benefit is how data can now be published in open, standardised formats and offers a way for data to be re-used by the public. This could be for example, customers of a business, suppliers of a business, or even a housing association with a need to access some government data for a study. With linked or open data and the centralised standardised way that data is published it is possible to build systems on this data as all changes can be dynamically reflected in the associated media that incorporates the data. (Bauer 2006)

Rowe (2011) indicates that more recently there has been a rise in the amount of raw government authority data that has been published on the web, this includes information on areas such as police crime statistics, prone flood areas, and geographical data etc, all with the aim of making information more readily accessible. Other government agencies, the research community, and the general public can consume this data in the knowledge that the data has been untouched, and use it in whichever way they choose, simply because of the way in which the data has been published,.

Linked data is not always necessarily open of course, and some confusion can usually be found in the terminology differences between linked data and linked open data. Both may follow similar frameworks for publication, but both are not necessarily open and accessible by everyone. Open data usually refers to the Open Data Initiative (ODI), a collaboration of different member organisations aimed at sharing and centralising open data. (Rowe 2011)

Standards 250w

In a publication by the IEEE written by Shadbolt, Hall & Lee (2006) the four main principles how linked data could be achieved are described. Tim Berners-Lee suggests the following as the standards to adhere to:

Use URIs to identify things.

Use HTTP URIs so that these things can be referred and de-referenced by both people and intelligent systems.

Publish machine-understandable information about the things when their URIs are de-referenced, using standard Semantic Web languages such as Web Ontology Language (OWL), Resource Description Framework (RDF), and XML.

Include links (with semantic annotations) to other related URI identified things in other places on the Web to improve discovery of related information.

By following the methodology above it ensures that the data being published complies with a standardised model and adds to the existing pool of standardised data. (Shadbolt, Hall & Lee 2006)

Taken from the Shadbolt, Hall & Lee’s paper, (2006) Appendix B illustrates the present makeup of the Internet; the new levels of abstraction were the inspiration of Tim Berners-Lee and are represented as 3 main layers these include:

Network Layer, as commonly termed, is concerned with the network of devices that are connected to the web.

Web Layer, is concerned with the web of documents

Graph Layer. The most recent addition which focuses largely on the area considered to be the Semantic Web. It enables users of the WWW or intelligent systems to create and explore connections between data objects without the awareness of boundaries.

Semantic Web Languages

RDF

OWL

SPARQL

3 Where is linked open data being used? 1000w

2.1 Participants in the ODI

2.2 Indicate growth pattern of ODI over the years and brief comment on new members

2.3 Current Projects

2.4 Emergency Services Look to Linked Data

2.5 QDOS, a much controversial area

Business choice for data i.e. Big Data

1.5 Legal, Ethical & Security Concerns? 300w

Comment on web browsing habit storage, and how it can be used against people

Comment on personal data that is already being exploited through web searching habits, i.e. products recently searched for constantly popping up thanks to google analytics

Comment on how an insurance company could use hereditary information, to base prices for customer’s premiums, or even settle claims or refuse to insure.

Heading more and more towards a big brother state

Commercially sensitive information could be linked my mistake

Results from awareness survey 700w

An objective account of what was found.

Some will be descriptive.

Some of this will be analytical DSS cross tabulations, averages, trends etc.

Conclusion 350w