Feb 17

VALA 2018 – Day Two – Wednesday 14 February – Data Day

And here are my notes from Day 2 – not including of course, the presentation that I gave with my manager Daniel Lewis.


VALA Conference 2018 – Wednesday 13 February

Plenary 3 – Linked Data Liminality – Matt Miller

Matt is a Metadata Librarian, programmer/developer, adjutant at a library school, worked in public and academic libraries and as a consultant.

Linked Data is an inevitable future, uses the web as a metadata platform, shared vocabularies and assign persistent identifiers to things.

Liminality is the in-between period, the threshold, the intermediate state. This perfectly describes linked data at this stage of its development. Its not there yet but it has gone beyond the start. This is a good reflection of society, which is always going through an intermediate stage of something.

Carnegie Libraries in the US and four in Australia (one in Northcote) were based around the building being funded by Carnegie, but the government in the area taking responsibility for stocking and ongoing running costs.

Linked data is a basically a list of things that are available so that you can find them. Our metadata is that list. But can we do more with that data than just make lists?

Beyond the record – expose the atomic level of data in the document where the document becomes the dataset, going beyond the metadata of the document; connect to external sources – connecting beyond systems to others – outside of libraries and other cultural organisations/

What does that nimbleness look like? Linked Jazz – documents to data and what happens next. The project involved oral histories transcripts of people involved in Jazz in the US. They developed an assisted automated system of marking names and connections and marking them up to the catalogue, this was also used to create a graph linking names in jazz. Also linked out to connect to external sources of information. Soon discovered that organisations in the sector were doing something similar so make it easy to connect the datasets, which can bring out new research opportunities and new ways to visualise data. WikiData has made it easier to do connections with Wikipedia entries.

Institutional Memory – how do you retain institutional memory, using linked data? NYPL has developed an amazing discovery system around famous authors – their works, works about them, their personal lives and more. Data.nypl.

Having discussions about linked data is also vitally important, as the data.nypl site has gone, due to a change in leadership. It needs to be embedded in your libraries work, to ensure it continues beyond a supporting manager/special funding etc.

Linked data as institutional memory need systems that support structured knowledge creation and storage – semantic annotation, semantic linking, integrating ad-hoc or orphan datasets.

Breaking out of the library – Harvard Library’s Case Law Collection. They digitised 40 million pages of case-law, converting it to XML as best they could, using OCR. But then what to do with it. They wanted to bring the data to where the people were, to build in more connections to facilitate discovery and use; and increase civic engagement, participation and education. They built a tool to facilitate this. Began by compiling a list of names of judges and then matching the names to the OCR content. The text is then put into the Wikimedia Commons system and ultimately into the Wikidata – both the original text and the metadata.

In this instance, where is the library in this, as once it is out there, we no longer have control over the data. Is this a concern for libraries or is it more important to get the content out there?

The only constant in this liminality of linked data is us, so we need to keep exploring, keep playing and questioning – and what better group of people to do that. (Us!)

Concurrent Session 7

Linked Data in Koha – David Cook Prosentient Systems

Completely self-taught in coding, although had some basics in his library course. He has learned that it OK that there are things that you don’t know.

Linked data = RDF + HTPP. RDF – subject, predicate and object and usually each part has a URI.

The Journey – best practices, tools, BIBFRAME, Experts? No real leaders, because organisations are using RDF experimentally in a range of different ways. From the engineering side he found software libraries, triplestores and then started finding software bugs with Koha, so fixed then and then found more and fixed them……

Koha is a MARC based system so had to work with the MARC records and then link the RDF to those records.

Linking data to non-linked data – use the URL of your OPAC – which theoretically is what should be in your MARC record. You don’t to import linked data, you want to directly link. The solution was to use the catalogue record as the subject and the you are in the linked data framework.

It still a work in progress, with David’s OAI-PMH and RDF support patches code working very well – but it is going through a rigorous assessment process by coders around the world. The code to display RDF in the OPAC is not yet complete and there is no system for dereferencing linked data, and there are so many more questions…….

The future: don’t know where the future is, but there is a lot of good work out there: Fedora Commons 4, Ex Libris Alma, Reasonable Graph, OCLC, New York New York, LIBRIS, Oslo Public Library, IT World, RDF, XMP and IIIF.

APO linked open data collections for public policy – Amanda Lawrence APO Analysis and Policy Observatory

Public Policy has a defined cycle of creation, but the actual cycle is a lot more chaotic. Public policy involves making decisions concerning society, but the sources and type of evidence that is used to make these decisions is prolific and often overlapping and sometimes clashing.

There is a spectrum of policy publishing – from formal and informal resources, including produced by publishing companies, produced by organisations and government and other which includes grey literature, ephemera, personal content and more.

APO has 40,000 records, 5000 organisations (not known as publishers) and 18,000 authors. They are used for their newsletters, hosted full text documents and data with full metadata, submissions are accepted. They are supported by partnerships, grand, community advertising projects and services, user contributed content, volunteers and linking collaborations.

How do they move to unstructured textual content to semantically structured and accessible content.

Linked data is another way of saying that data is being cleaned up to make it more usable.

They have funding for the ARC LIEFT Project 2019 – Linked Semantic Platforms for social, digital and inclusion policy. Project RoadMap is discovery & analysis, inter-operability, connected collections, collection methods and content. Looking at sourcing content and metadata using multiple tools and approaches.

APO is already interoperable with many organisations. You may find that your content is being utilized by other organisations already without linked data – it is not the only solution.

In 2018, they will focus on tools for discovery – pulling data together from disparate sources and being able to present it using visualization. Will be demo-ing a pilot project with Research Graph at RDA Berlin 2018.

Open systems and the innovation opportunity – Neil Block EBSCO Information Service

Around 15 years ago, the library systems were legacy, had limited functionality, limited interoperability but the market was robust.

In the Integrated era, functionality went up, interoperability started increasing, but choices started to be limited. Discovery and linking also appeared in this time.

Next Generation Era, functionality is going sideways, interoperability is affected and there are even fewer options.

FOLIO is a global community supporting the creation of a new open source Library Services Platform (LSP). FOLIO will includes LMS functionality and is designed to work with other apps.

The group aims to build a dynamic, sustainable community; deliver an open software platform to encourage innovation; utilize open source to drive down costs; encourage choice of apps with microservices architecture, open innovation model is distributed, more participatory and centralized.

Products have features, platforms have communities. Van Alstyne, Parker rand Choudary – Harvard Business Review.

With commercial LSP – direction is controlled by the company, 2-sided relationship, hosting and support from one vendor, user groups and enhancement lists not focusing on your needs, delivered as is. Platform (open source) – directions supported by community, multi-sided relationships, hosting and support options, multiple ways to add enhancements, foundational technology: built to encourage innovation.

FOLIO began 2 years ago with three foundation partners – Open Library Environment, EBSCO and Index Data. More than 20 partners including University of Sydney, University of Newcastle, JISC.

Innovation through collaboration: 9 different teams. They are developing micro-networks that are coding for particular areas that can be used, reused and discarded for better code if it arises. 60 programmers, 150 librarians and other specialists are involved.

All decisions are made using open source channels, including Slack.

FOLIO Ideation was done old school (whiteboard and post it notes), librarians and developers worked together to talk UX and develop ideas. FOLIO is being based on the prototypes that are developed and approved by these reference groups.

FOLIO could be used as an LMS, or as a way of collecting and presenting content from libraries using disparate LMS’s.

Libraries are investors in FOLIO – bringing their knowledge and their experience.

FOLIO Platform Benefits – built around an open community, supports local innovation, provides choice.

FOLIO Challenge – looking for ideas. They also offer grants to developers.

Concurrent Session 12

Daniel and my presentation.  See it at: https://www.vala.org.au/vala2018-proceedings/1241-vala2018-session-12-lewis

Translating disruption into action: next steps for a 21st century library service – Natalia Fibrich Library Training Services Australia

The period we live in is known as the 4th Industrial Revolution. It is characterized by breakthroughs in technology. But there are also changes in society – including how we interact, how we live and the infrastructure around us.

Disruption si about stopping something from continuing as expected. It can be a force for good, to change things to be better versions of themselves. There are a number of traps that we need to avoid: physical trap – investing in old systems instead of new; psychological – past successes rather than new options: strategic trap – when the focus is on today and not on the future.

It is important to note that disruption happens to people.

Translating disruption into action: a model for organizational disruption.

Context – if there is no roadmap, you go nowhere. The library needs to understand the megatrends that impact the library and community; trends impacting the LIS sector and the local community; Issues – that impact the library in short to middle term and internal context. It is not possible to mitigate them all, but it is important to identify and consider them to help the process get started.

Leadership – many are not prepared for the challenges of change. Good leaders recognize that their organisation is every changing. They are redesigning their organisation’s ‘business-as-usual’ on an ongoing basis. The best way a leader can facilitate change is to walk the talk.

Team – organisation’s need to invest in their staff. Top 10 skills needed for the future: sense-making, social intelligence, novel and adaptive thinking, cross-cultural competency, computational thinking, new media literacy, transdisciplinarity, design mindset, cognitive mind management and virtual collaboration.

Customer – Libraries need to ensure they are customer centric – which includes both current users and potential customers, through the many ways that they can interact with the library. Every form of contact should be considered to ensure that it is indeed customer centric.

Organisational culture – in the current environment, our organisations are transparent – the public can see an organisation’s culture. It is therefore not only important for staff, but also for our customers. This culture exists, whether it is deliberately or unintentionally created. To help create a positive culture you need a vision, mission and values which are not just words, which everyone lives. Another challenge is efficiency – where we miss the opportunity to innovate because we are further streamlining.

Organisational strategy and design:

Organisational change is an ongoing Work in Progress.

Sari Feldman – quote “While these disruptions may feel like a sea change in libraries, our work remains grounded in an enduring ideal: People walk through our doors with ideas, ambitions, and challenges, and we meet them with resources that foster individual opportunity, options, and optimism.

Libraries value today is more what we do and for people.

The revolution will not be standardised – Angela Galvan

Puppets are like technology – machines requiring human involvement to tell a story.

Libraries are not and have never been neutral – we have inherent biases in all aspects of libraries, including our collections, our systems (which are created by people and reflect their biases). Libraries are living things because of the people working there.

Get out of the way – theirs and ours. There is a difference between ultimate mastery and foundational knowledge which allows you to have a conversation. (Chatted with someone next to me about where we are stuck).

Doesn’t work, buy it. We can stop supporting mediocrity with free labour today. We assign values and we are assigned. We said we wanted some things and so we did it, and we have kept on saying we want it. But do we really want it? Or maybe we did, but we don’t anymore

Magic – disappeared costs. Many of us are told what we do is magic. But its only magic if the way it works remains unexposed. Some of our processes revolve around what library staff require, not what is best for the user.

Glitches – are the unintentional exposure of values. When libraries link to websites, we give it a little more authority. (paraphrased Jessamyn West). Google can give authority and relevance to harmful content. Library jobs are about connecting people to information, vendors are about making money. It’s an agenda that competes with mine. AI is concerning as people are developing the technology without considering history or ethics – the developers are building their biases into their creations.

Resistance and compassion – we need to ensure that there is compassion in all our systems and processes, including discovery.

We love our work and we love helping our users, but they will never know who we are and that is the way it should be.