My Blog

Linked Data 2017


Linked Data and other Semantic Web technologies are used in a variety of
organisations for projects large and small. They are a key feature in both
government open data initiatives and enterprise-wide data integration systems
for internal use. In this course, you’ll learn about the building blocks of the
Semantic Web and how to use them, including how to model your data in RDF,
integrate with third-party and OpenData sources, and how to enter and run SPARQL
queries. You’ll hear about how these technologies are used today, and have a
chance to try them out in the hands-on portions of the classes.

This course is chaired by Kal Ahmed and taught by Andy Seaborne, Cathy Dolbear, Claire Hainsworth, Jen Williams, Kal Ahmed, and Stuart Williams.

Classes for 2017

The Linked Data course runs on

An Introduction to Linked Data

Taught by Jen Williams.

Linked Data is a set of best practises that defines how to structure data so that
it can be interlinked and published on the Web. By using the Web as the
publishing platform, data can distributed across innumerate points, while still
allowing navigation and discovery through the use of URIs in the data itself.
Data models are embedded into the structure of the Linked Data itself, and
follow the same best practises by being defined as Linked Data Vocabularies.

The principles of Linked Data allows developers to publish data without the need
to develop custom APIs, and offers the ability for data publishers to adopt data
standards to improve interoperability between disparate sources of data.

This “Introduction to Linked Data” course will cover all of the basics around the
structure of the data itself, the use of URIs, popular vocabularies and how to
use them in sample datasets.

Delivering Linked Data

Taught by Kal Ahmed.

The infrastructure of the world wide web can do more than deliver documents for
people to read off of their screens: it can also deliver data for applications
to use. The principles of Linked Data have laid a foundation that has made it
possible for governments, media, and e-commerce retailers to publish data on the
web without depending on custom-built APIs. This class will show you how to take
advantage of these principles to consume available data and to publish it

Among other things, we’ll learn about popular sets of linked data that you can
use, how to create links between datasets, how to mint good URIs, HTTP issues,
and how to take gradual steps toward good linked data publishing.

Introduction to SPARQL

Taught by Andy Seaborne.

SPARQL is the standard W3C query language for semantic web applications. It
brings together the features of a number of RDF query languages into one method
for extracting information from data represented in RDF, whether small datasets
or large.

This session will provide a solid grounding in SPARQL. After demonstrating how
powerful some very simple SPARQL queries can be, we will take a practical
approach to looking at the key features of SPARQL 1.0 and 1.1, and then explore
the principles underpinning the SPARQL query language.

Hands-on SPARQL

Taught by Andy Seaborne,
Jen Williams,
and Kal Ahmed.

In this session students will get a chance to try their hand at writing some
SPARQL queries. This session will help to consolidate the fundamentals of SPARQL
query as well as introducing some new tips and tricks for being more effective
with the SPARQL that you write.

For this session students will need a laptop with a wi-fi connection and a
browser capable of running modern Javascript applications (e.g. Chrome or a
recent version of Firefox).

From CSV to RDF

Taught by Kal Ahmed.

CSV (comma-separated values) files are one of the simplest ways we have of
sharing data. From databases and spreadsheets to content management systems and
CRMs, many applications are capable of producing CSV data.

This session will focus on how we can create and publish Linked Data from this
popular data format. Using the free publishing service we will show
how easy it is to map simple CSV data to an ontology and publish it. We will
then go on to show how to make use of common tools and open-source products such
as OpenRefine to get more creative with formatting and structuring data for

RDF Modelling

Taught by Stuart Williams.

Modelling the world in RDF involves identifying the important entity types within
some domain of interest, their properties and relationships. In this session we
will workshop the development of an RDF data model for a familiar domain.

We’ll look at the open world nature of RDF models and their focus on making
statements about things in the world. We’ll build up a data model as we go
along, questioning its purpose and considering the questions (possibly in the
form of SPARQL queries) we’d like to be able to answer from data expressed using
the model. We’ll take a quick look at expressing our evolving model in RDFS and
OWL. Finally we’ll contrast the model we develop in the session with one that’s
already been developed for the same domain.

Linked Data In Practice

Taught by Cathy Dolbear and
Claire Hainsworth.

In this session, guest speakers from industry will give brief talks on how they
are using Linked Data in practice.

Experiments in Using Linked Data to verify Knowledge Graphs in

This talk presents the use case at Oxford University Press of using knowledge
graphs for named entity recognition, and how we are trialling the use of Linked
Open Data sources such as DBpedia and Geonames to verify and disambiguate the
knowledge extracted from our content data. Cathy will describe how we can move
from simple subject categorisation to a richer, more nuanced description of the
people, places, organisations and other things mentioned in a text, based on a
separation of concerns between the content XML, the RDF metadata layer, and the
entity knowledge graph.

Linked Data at the Environment Agency

This talk presents the use of Linked Data at the Environment Agency, an
environmental regulator that works with businesses and communities to protect
and improve the environment, for people and wildlife.

Linked Data Quiz and Panel Session

Taught by Andy Seaborne,
Cathy Dolbear,
Claire Hainsworth,
Jen Williams,
Kal Ahmed,
and Stuart Williams.

Test what you have learnt with this fun quick-fire multiple-choice battle! Points
are scored for being the fastest to pick the correct answer, but beware
sometimes there is more than one right choice!

To round off the Linked Data track, all of our presenters and guest speakers will
form a panel to discuss the practical aspects to publishing and consuming Linked
Data. This year the focus is “Collaborating and Competing”. The discussion will
focus on the question of how publishers can collaborate on publishing Linked
Data while still competing in an open marketplace.

This is an opportunity to participate in a lively discussion and to put your own
questions to the panel and draw on their collective wisdom and experience.