Resource discovery priorities and session synopses

Resource discovery priorities

The sessions from Discovery 2013 came up with a list of 18 priorities relating to resource discovery from the perspectives of users, technology and progress.

The top six priorities were discussed in the panel session.

Please note that there will be some work done to refine some of the below priorities to make sure that they are realistic and measurable.

Widen the view of what we mean by metadata generation and contextual metadata
Develop use cases for metadata share evidence and embed best practice
Empower the teacher and humanise the presentation of information and increase contact time through the online experience
Initiative to promote skills and working together with researchers; work to understand role of researchers
Initiative to aggregate and distribute skills, knowledge and expertise across the sector - librarians, curators, archivists and developers
Engage and understand end users (and how they add value)
Create a national brand to communicate as an aggregate voice
Identify and engage with 10 new user communities as innovators and business
Seek consensus that technical, legal and semantic interoperability is a must.
Segment and analyse the student market and develop an engagement plan for each segment
Machine access to controlled vocabularies in graphs
Demonstrate value of access to data through explotation of indexes, APIs and emergent visualisation technologies which users can use
Renegotiate success criteria with funders so that we can ensure continuation of infrastructure investment as that infrastructure begins to vanish
Identify unique data features and common resuable data
Given a rapidly developing environment, work with the community to understand and exress the rationale and use cases for aggregations
Transition existing aggregations from 'take and publish' models to enable contributor publication and aggregation to add value
Develop the habit of anticipating user expectations
Find our sweet spots and define a minimal framework for working together

Session synopses

Theme 1 - Are we failing users? Can open approaches meet their needs?

Abstract

All the work on resource discovery that we are aware of is driven by the desire to improve the experience for end users. To make it easy to find and explore content, and to locate and access it once it has been discovered. Many of the approaches around the world have used open data approaches to achieve this aim. Are these open approaches working?

Of course, users is a very broad term so we want to use the parallel sessions to focus on some of the most important user groups:

Researchers rely on library museum and archive content to do their work. How can we improve the services that we offer to them to ensure that they can find and access all of the content that may be relevant to them?
Students, undergraduates, postgraduates, on campus and off, all rely on resources to complete their studies. How can we ensure that they can find and use these resources using whichever services or device they prefer?
Library, museum and archive resources are also used outside the educational domain to support innovators, the general public and commercial enterprises. Are we serving their needs?

Plenary

Maura Marx, Director, Digital Public Library of America Secretariat.

Parallel session 1- Where are we failing researchers and what can we do about it?

Main issues discussed

There need to be different strategies for non-digital and digital discovery approaches.
How to involve researchers in metadata generation.
Linkages. Discovery isn’t just about catalogues. It’s not just search and discovery. We’re missing browse search and synthetic results.

Actions

Initiative to promote skills and learning together with researchers + work to understand role of researchers.
Widen the view of what we mean by metadata generation and contextual metadata.

Parallel session 2 - How can we improve the student experience?

Main issues discussed

Be where they are: it used to be that we built workflows round us, when resources were scarce and we were a primary resource but that’s no longer true, so now we need to build the library round the user and recognise that the user has many resources and start in very different places than 5 years ago. eg undergraduates go first to Google, then to ask other people, even Facebook and text. We need to interface with students where they start their search and approach what we do from the student point of view, take into account the student experience.
Segmentation: students are not just a homogenous group.
Undergraduates with three years ahead of them use resources in very different ways to postgrads, research students and so on, and have much less knowledge of the tools and resources. Students also study differently in different contexts, not only based on why they’re there and what they need on a particular occasion but also on where they are culturally and geographically. We need to segment the market and respond differently to different student categories and different contexts.
Human guidance: students are not equipped to evaluate the quality of what they find, which can be voluminous. We need to help them evaluate and understand authority, and also humanise the sterility of data, giving guidance at the point of use, which might be real-time (eg a chat box that appears if a student gets a ‘0’ return on a search) or might provide human context (eg recordings of David Attenborough or AJP Taylor).

Actions

Segment and analyse the student market, and develop an engagement plan for each segment.
Empower the teacher and humanise the presentation of information and increase contact time through the online experience.

Parallel session 3 - What do we need to be doing for innovation, society and business?

Main issues discussed

It’s no longer about metadata or cataloguing, it’s about full text search and google and we need to be engaging with these commercial providers - google, facebook, Elsevier - because that’s where our users start their search. Users also need educating in how to use these sophisticated search tools.
Non-commercial licensing - and licensing in general - needs to be clearer and institutions need to make it clear that opening up is part of the creative economy as a whole, to make the business case and think as a sector not just as an individual organisation.
The sector needs to develop an aggregate voice and identity - a national infrastructure to sustain the projects we create. If the DPLA can do it in a country as complex as the US, then we can do it here.
Developers need to feel confident that data that is produced by organisations eg as APIs will be sustained, at least for a certain amount of time. It’s about sustainability, communication, engagement, social responsibility - publishers of any kind of open data have a responsibility to the groups that use it. They need to say what their own intentions are, put dates on data and reassure, where possible, that things are stable. They also need to engage with what happens next.

Actions

Identify 10 user communities to engage with as innovators and business.
Create a national brand to communicate as an aggregate voice.

Theme 2 - Technical coping strategies for resource discovery

Abstract

This theme focuses on the technical issues faced by libraries, museums and archives in making sure that people can find and use their resources. We have chosen 3 issues that we have wrestled with in the UK to explore in the parallel sessions:

There is a wealth of emerging technologies and standards that libraries, museums and archives could use to make their resources discoverable. Which should we concentrate on?
Aggregation of content and metadata has been a core part of efforts to improve discovery of resources for many years. Does it still make sense for libraries, museums and archives? If it does, how should we approach it given what we have learned about aggregation from past endeavours?
The issue of data quality has raised its head again and again throughout our work to improve resource discovery in the UK. We would like to try and use this discussion to identify where data quality is absolutely essential and stops us from delivering the services that are required by end users. We would also like to think about how we can address those scenarios where data quality stops us delivering services.

Plenary

Paul Walk, Director, Innovation Support Centre, UKOLN

Parallel session 1 - Which emerging technologies do we need to focus on in 2013?

Main issues discussed

Battlegrounds: The potential “battleground” between consumers and content suppliers - users are bringing their own devices, and infrastructure (eg things they have on the cloud)... Content suppliers risk falling behind. Is there a way to exploit the rich personal context people bring in? The battleground between ontology and linked data - and need to get the two working together. This combined with a discussion of the (for eg) data loss problems when things are pushed into large aggregators.
How to combine “machine learning” (e.g. advanced search methods through ontologies and data, greater semantic richness), and the exploitation of networks and human intelligence.
How to make the most of “context awareness”.

Actions

Machine access to controlled vocabularies in graphs.
Demonstrate value of access to data through exploitation of indexes, APIs and emergent visualisation technologies which users can use.

Parallel session 2 - Do we still need to aggregate?

Main issues discussed

Yes, but aggregate what: content or metadata? This should not be about creating web portals but creating resources for other developers to do useful things for their communities, offering APIs and documentation to enable developers to understand the why, the what and the how
Are aggregations effective? And isn’t everything an aggregation? And don’t our users use Google, Facebook etc? So should we not just leave aggregation to Google etc and work at helping people to publish/expose their data.
But why do we aggregate? Is it to increase audiences, to provide resources for specific projects or to enable users to discover what we academic institutions hold. Google doesn’t do that and existing aggregators are imperfect: there are too many, none holds ‘all’ the data and users don’t know which to choose. We need to understand why aggregations exist learn from them for the future, to we change the way we aggregate in order to suit what customers need.

Actions

Given a rapidly developing environment, work with the community to understand and express the rationale and use cases for aggregating.
Transition existing aggregations from “take and publish” models to enable contributor publication and aggregation, to add value.

Parallel session 3 - What does data quality stop us doing?

Main issues discussed

Utility - is the usability of the data where quality lies? It may be that it is in its aggregation or accumulation that the quality emerges. Quality is inherent in the usage of the data itself. But, for developers, when you put content in an aggregator site it's mapped to a common scheme and sometimes the quality - what really makes it interesting – is lost when it becomes squished into a common pot and some fields are lost.
End users – are they the best judge of quality? Less is not always more, sometimes less is less. Audiences recognise quality when they see it and bad quality stops us from developing better quality. Look at your organisation's website and if you think that the quality is low then be sure that your audience will also think it's rubbish. Bits of your data will also be seen out of the context you are used to – understand what it will look like on the ground, look beyond your own horizon.
Where do interventions need to happen? There are examples of people sharing their lessons. Who hears about that and who doesn't and how to make it discoverable to them? Certain amount of hard-won knowledge are out there already, needs to be disseminated further. Same with use cases - these are really needed.

Actions

Develop use cases for metadata, share evidence and embed best practices.
Identify unique data features and common reusable data

Theme 3 - Exploration of effective progress

Abstract

There are a lot of us working on improving resource discovery around the world. We all have different focuses and different priorities but what is striking is that there are so many similarities in approach between all these initiatives. Most are encouraging openly licensed metadata and, where possible, content. Most are using technical approaches based on web best practice. Many are exploring linked data. Nearly all are trying to engage developers to work with the content to develop innovative new approaches to resource discovery.
These similarities seem to present an opportunity for working closely together. How do we seize that opportunity? These parallel sessions are designed to explore aspects of this big question and to identify concrete actions that can be taken to make sure we make the most of our similarities.

Plenary

Alastair Dunning, Programme Manager, European Library

Parallel session 1 - What’s going to disrupt us and how will we react?

Main issues discussed

Changing funding structures. Related issues were developer flight and maintaining funding after data suppliers have made an infrastructure that’s invisible.
What to do when organisations we rely on (google, in other words) suddenly change their criteria for e.g. search. I.e. How to cope with changes made by third party organisations we depend on, but over whom we have no influence.
How to get the benefits from Open Access. A point raised was that “OA is costing us money, but are we getting the benefits?”
How to keep up with changing user expectations.

Actions

Renegotiate success criteria with funders so that we can ensure continuation of infrastructure investment as that infrastructure begins to vanish.
Develop the habit of anticipating user expectations.

Parallel session 2 - What changes do libraries, museums and archives need to make to support better resource discovery?

Main issues discussed

We need to revisit the idea of "put it out there and it'll happen" and be strategic about putting it where users are.
Users are not a single audience. We need to segment the audience and focus on the segments that best answer out business case.
Users can also be creators and data sources, and vice versa.
Creators may be the best source of metadata and usage data can add value to metadata.
We have an obligation to provide metadata. Otherwise it's like a business selling products without a proper description.
Discovery isn't the problem: access is the problem. This isn't just about Finch but about access to non-digital resources and how to provide the right information for users to choose how and whether to access them.
There's a need to bring our different communities together so that non-technical people overcome the jargon and developers understand their needs.

Actions

Initiative to aggregate and distribute skills, knowledge and expertise across the sector - librarians, curators, archivists and developers.
Engage and understand end users (and how they add value).

Parallel session 3 - There are a lot of us working in this area. How do we work effectively together?

Main issues discussed

Definitions and context. Need to first of all understand who is "we" and what we mean by partnering. What is the common vision and shared principles? What are the standards we want to adhere to and comply with? What are we committed to? We need to understand what we are trying to get out of the relationship and where the "sweet spots" and "pain points" are.
International perspective. From a sustainability viewpoint it's got to go beyond a European and American perspective, got to look at China and India. Pretty soon any data project becomes global. There is only one world wide web. There is only one planet. We have a collective opportunity to improve the world wide web and to do that you have to use the whole web not just one website – it won't happen on Europeana and DPLA and all the other locally focused sites. To change the world you need to think differently.
What would success look like? It looks like when you're an end user and our "stuff" coheres in a natural and painless way around you to help deliver a functional and literate society. The web is a machine for anything and has a wonderful way of working around blockades to knowledge. This is about creating a literate, functional and enlightened society.
An obvious sweet spot for collaborative working is in the area of vocabularies, this is something intended to be shared and where more effective linking and connections can be made.

Actions

Find our sweet spots and define a minimal framework for working together.
Seek consensus that technical, legal and semantic interoperability is a must.