Session 5: Unlocking Data: Exploring Different Roads
Wednesday 26th June 2019 – 14:45 – 16:15
Chair: Matias Frosterus, National Library of Finland, Helsinki, Finland
5.1 Open Knowledge Maps: a Visual Interface to the World’s Scientific Knowledge
Peter Kraker, Open Knowledge Maps, Austria, Najmeh Shaghaei, University Library of Southern Denmark, Denmark
Getting an overview of a research field and being able to identify a set of relevant findings pertaining to one’s information need are prerequisites for research, evidence-based practice and self-directed learning alike. Yet, the tools for exploring and discovering scientific content are often lacking. With traditional, list-based search engines, users have to examine articles and their relationships by hand, which is a time-consuming process.
Open Knowledge Maps is an attempt to transform discovery of scientific knowledge by providing an open, community-driven non-profit system that leverages the digital open science ecosystem. Instead of lists, we propose to use knowledge maps for discovery. Knowledge maps provide an instant overview of a field by showing the main areas of the field at a glance, and papers related to each area. This makes it possible to easily identify useful, pertinent information.
Our goal is to provide a large-scale system of open, interactive and interlinked knowledge maps for every research topic, every field and every discipline. This system will enable people inside and outside of academia to not only get an overview of a field and identify relevant concepts, but also to discover trends, recognize important researchers, and to understand connections between fields.
On our website https://openknowledgemaps.org, users can currently create a knowledge map for a topic of their choice based on either BASE or PubMed. Our software retrieves the 100 most relevant results for a topic and generates a knowledge map based on textual similarity between the records. The map is intended to give users a head start in their literature research.
With this service, we have created a lot of enthusiasm in the community. Our user base has quickly grown: since our launch in May 2016, we have recorded over half a million visits to the site and more than 120,000 maps have been created. Open Knowledge Maps has become an international collaboration with team members, advisors and partners from variety of fields, including research, librarianship, design, software development, citizen science, and the open knowledge and open science movement.
In the future, we want to turn discovery into an open and collaborative process. Most people are currently tackling discovery on their own – and therefore repeat the same process over and over again. By sharing the results of our discoveries, we can save valuable time and build on top of each other’s knowledge; for example, researchers and medical librarians can collaboratively map the newest research on a certain disease and openly share result of their efforts for the benefit of evidence-based practice and patients affected by this disease.
We see research libraries and librarians as central to this vision. The complex collaborative system outlined above cannot be realized without experts on knowledge stewardship and community engagement. Together with the other stakeholders from research and society, including researchers, students, journalists, citizen scientists and many more, we want to create system that enables us to create pathways through science for each other. So that we can all benefit from this unique knowledge.
Dr. Peter Kraker is the founder and chairman of Open Knowledge Maps, a charitable non-profit dedicated to dramatically increasing the visibility of research findings for science and society alike. To this end, Open Knowledge Maps operates the world’s largest visual search engine for scientific knowledge. Prior to founding Open Knowledge Maps, Peter was a senior researcher at KnowCenter, Austria’s leading research center for data-driven business and big data analytics, managing the topic of Open Science. He was involved a number of EUfunded projects, most recently leading the work on innovative dissemination in the Horizon 2020 project OpenUP. Peter coordinates the GO FAIR Implementation Network Discovery and is a core team member of the Open Science Network Austria (OANA). A long-time open science advocate, he is known for coining the term Open Methodology and for his leading role in creating The Vienna Principles – A Vision for Scholarly Communication in the 21st Century.
Dr. Najmeh Shaghaei holds a Ph.D in Business Management and works as Head of Library – Campus Sønderborg and Internationalization Project Leader at the University Library of Southern Denmark. Shaghaei has a background as a researcher in Change Management and Knowledge Management and is currently working on a number of international projects and researches related to Knowledge Management and Open Science. Dr. Shaghaei is a member of LIBER Leadership Program working group, also working with Open Knowledge Maps organization in Austria as Community Coordinator, and moreover member of Peer Review Board of Turkish Librarianship Journal.
5.2 NAISC: A Linked Data Interlinking Framework for Information Professionals
Lucy McKenna, Christophe Debruyne, Declan O’Sullivan, ADAPT Centre Trinity College Dublin, Ireland
Publishing records as Linked Data (LD) has many potential benefits for Libraries including exposing collections to a larger audience and allowing for more efficient user searches. Despite this, relatively few libraries have invested in LD projects and the majority of existing projects display limited interlinking of resources across institutional datasets. As one of the fundamental prerequisites of the SW is the existence of large amounts of meaningfully interlinked data, it is key that libraries not only publish their records as LD, but that they also interlink these records with other authoritative datasets on the SW. Library datasets that have been enriched with interlinks would allow for seamless navigation between internal and external datasets, guiding users to a web of related information from a single information search.
A survey was conducted to understand Information Professionals’ (IPs) position with regards to LD. The survey was completed by a total of 185 international librarians, archivists, metadata cataloguers and researchers. Results indicated that IPs find many current LD tools to be technologically complex and unsuitable for the needs of the library domain. Results also showed that, when attempting to interlink LD resources, IPs find the process of ontology and property selection to be particularly challenging.
Selecting an appropriate property, or link term, to interlink a pair of resources is a known challenge across the SW, not just within the library domain. Our research is focused on identifying the types of interlinks IPs wish create to between two resources, how to accurately describe these relationships in LD, and how to describe the decision making processes behind creating these interlinks. This has led to the development of NAISC – an interlinking framework and tool designed to facilitate increased IP engagement in the interlinking process. NAISC stands for the Novel Authoritative Interlinking of Schema and Concepts, and is also the Gaelic word for links. NAISC includes a set of processes and standards for the creation of an interlink between an internal and external resource. NAISC also includes a provenance data model that allows for the description of the decision-making processes behind the interlinks generated. An accompanying graphical user interface, that guides IPs through the NAISC interlinking process, has also been designed.
We adopted a User-Centred Design Science Approach to developing NAISC to ensure that it meets the unique requirements of IPs. As per this approach, each version of NAISC has been evaluated by IPs, with the results of each evaluation being used to iteratively refine the framework.
At LIBER 2019 we will discuss the results of our Library LD Survey and the design process behind the development of NAISC. Additionally, we will demonstrate the most current iteration of NAISC and discuss how NAISC could be applied by libraries to interlink their LD datasets. We will also focus our discussion on the process of link term selection and the provision of provenance data as a means of publishing more authoritative interlinks.
Lucy McKenna graduated with a Masters in Library and Information Studies from University College Dublin in 2016. Lucy is now in her final year of study towards a PhD in computer science at the ADAPT Centre in Trinity College Dublin. The ADAPT Centre, funded by Science Foundation Ireland, is a dynamic research centre based across four Dublin Universities: Trinity College Dublin, Dublin City University, University College Dublin and Dublin Institute of Technology.
5.3 Mining for Europe
Alex Fenlon, University of Birmingham, United Kingdom
With the Directive on Copyright in the Digital Single Market going through the Trilogue stage of the legislative process, there is a very real chance of an EU wide text and data mining exception coming into force in the next few years. The precise nature and nuance of the exception are still subject to debate but the prospect of an exception, even one limited to research institutions and non-commercial research purposes, is exciting. It offers huge potential for EU lead developments in 21st century technologies, and opens up new and tantalising research opportunities.
Many e-resource licences agreed by libraries allow institutions to archive and preserve content licensed during subscription periods. Services such as Portico, Lockss and Clockss all contain a huge wealth of journal, and similar content, structured, organised and ready should suppliers and publishers withdraw from the market place, or face technological issues.
There are some differences but it seems that the UK exception is very similar to the one being discussed by the EU. It is likely that researchers engaged in non-commercial research will have the ability extract knowledge from data sources without the need to license the content from owners. This ability will apply to any source of data that the researcher has legal access to including online sources openly available, and content available under licence via an institutional library for example.
The prospect of an EU wide exception for text and data mining combined with the opening up of access to ‘dark’ archives and/ or institutional repositories represents a fantastic opportunity for researchers. Being able to access and acquire content from across the EU without the need for extensive complex licensing and legal support, offers a competitive advantage for EU researchers and institutions over and above many of their international peers. A combined platform with data, knowledge and information from across the archives and repositories could lead to more world leading research arising from EU institutions impacting on millions of individuals.
Presenting on behalf of Alex Fenlon:
John Dowd is the Assistant Director for Collection Management and Development at the University of Birmingham. John’s team oversees the library print and digital collection as well as the Scholarly Communications and Copyright and Licensing teams. Given the shift and increase in the deployment and exploitation of digital tools to support research activities John’s teams are central to researchers not only in enabling access to content but also in ensuring it is fit for research purposes. John co-chairs the Midlands University Academic Libraries (MUAL) purchasing consortium.