2020 LD4 Conference on Linked Data in Libraries

Session recording (Picknally Camden & Hahn; Khan & Usong)

Prototyping and Evaluating the Share-VDE (Virtual Discovery Environment) Linked Data Discovery Interface, Beth Picknally Camden and Jim Hahn (University of Pennsylvania)
The Share-VDE (Virtual Discovery Environment) has engaged a global partnership, with library-driven development.The project encompasses enrichment and conversion from MARC to BIBFRAME/RDF, creation of a cluster knowledge base, development of manual and automated tools for interacting with data, and the creation of a linked data discovery environment. University of Pennsylvania Libraries have been a Share-VDE partner since its inception in 2016, and, in 2019, embarked on a special project with Casalini, @Cult and Samhaeng for further development on the user interface. Penn’s goals are to demonstrate discovery in a linked-data user environment, and to enhance usability with APIs that will allow users to find resources and request delivery (local or ILL). This presentation will focus on the process for developing the prototype (UX/UI design, mock-up, technology review, etc.) and include a demo of the prototype. Additionally, we will articulate a set of interface evaluation metrics in the form of a heuristic research agenda for linked data discovery. The research provides an understanding of the capabilities of the Share-VDE interface to support the user tasks promulgated in the IFLA Library Reference Model (LRM).

The library catalog and linked data: a tale of two technologies, Huda Khan (Cornell University), Astrid Usong (Stanford University)
In the Linked Data For Production: Pathway to Implementation (LD4P2) project, we employed a user-centered approach to explore the integration of linked data sources in library discovery interfaces Through a combination of user interviews and evaluations of mockups and prototypes, we sought to better understand how linked data can meet user needs around search in the library catalog that extend beyond known-item search to more open-ended discovery tasks. We have also investigated concrete approaches for using Schema.org to improve the indexing of our catalog records by search engines. In this session, we review our design and prototyping efforts, Blacklight integration work, and the user research and evaluations we have conducted to address the four main discovery areas laid out in the LD4P2 grant: knowledge panels, browsing, semantic search, and the use of Schema.org. We will discuss the larger questions of intended user experience and delve into what we learned from users about our mockups and prototypes incorporating context, suggestions, and relationships to subjects, people, and collections. We will also discuss specific challenges and lessons learned while developing Blacklight prototypes bringing in data from both linked data sources such as Library of Congress name and subject authorities, Wikidata, DbPedia, FAST, and VIAF; as well as from external sources such as Who's on First, Discogs, Google Books search, the Open Syllabus Project, and Cornell Digital Collections. Some challenges for integrating data from non-catalog sources include inconsistent availability of connections between catalog entities and linked data sources, consistency and comprehensiveness of the data, reliability, and the performance of queries or data retrieval. We will also discuss possible future directions for improving user experience and integrating this work into Blacklight.

Book actions and Linked Bibliographic Data at Google, Erin Dobias (Google) (This presentation will not be recorded.)
Over the last few years Google has worked on exposing library holdings for ebooks, audiobooks, and print books in both Knowledge Panels and on the new books.google.com, allowing users to explore the holdings at their local library and discover formats that they may not have previously considered. In this presentation, I will discuss how this data is exposed, and how the increased exposure has driven the need for Work IDs, work/manifestations relationships, and manifestation/manifestation relationships in bibliographic data.

Session recording

Where Linked Data Can Impact User Experience: Findings from a Study on Discovery Workflows, Emma Boettcher (University of Chicago)
Slides available: https://static.sched.com/hosted_files/ld42020/e0/2020LD4_discovery_boettcher.pdf
In 2019, the University of Chicago Library conducted interviews with 19 students and faculty members about how they discover literature for their research. During the sessions, the interviewers did not mention linked data once – and that was intentional. Unlike usability studies that put a solution in front of the participant and evaluate whether it solves their problems (or produces new ones), this form of user research allowed us to learn more about existing workflows and challenges in our participants' research. With each interview, we learned what relationships between resources our users value during discovery, how they measure the usefulness of search results, and where they see themselves situated in their fields. This presentation will go over the findings from this study, focusing especially on the ones that have implications for linked data for discovery. It will cover how library users seek out related resources and what relationships matter to them; how date, author, and other parts of an article are used and evaluated for relevance to the researcher's topic; and the challenges encountered by newcomers to a topic, even if they consider themselves experienced researchers in other fields. Based on these findings, the sessions will also suggest where future work, both in user research and in linked data for discovery interfaces, could make the most impact. While the presentation may be of most interest to UX researchers and designers, it should also be of interest for anyone who works in discovery at their library.

A research agenda for the evaluation of semantic search interfaces, Jim Hahn and John Mark Ockerbloom (University of Pennsylvania)
Slides available: https://works.bepress.com/john_mark_ockerbloom/20/ 
For research libraries to move successfully from experimentation to implementation with library linked data and semantic search interfaces, we need to better understand how these systems can best support users' information discovery needs. This presentation outlines a research agenda and methodologies we will use in the University of Pennsylvania Libraries to evaluate, implement, and extend library discovery systems using enriched and linked metadata, including systems used in the LD4 community. Our intended audience is librarians and developers of linked data systems with an interest in discovery. Our research agenda begins with user tasks described in the IFLA Library Reference Model (LRM), but also considers extensions to those basic tasks that linked data-enabled systems support. These extensions include enhanced topical browsing; the discovery of works, people, and topics across multiple information collections; and the selection and delivery of the most appropriate copy of a sought-after work in a multi-library context that takes into account both content and obtainability. Our evaluation methodologies include both quantitative and qualitative analyses. We will begin with analyses of current discovery search logs, and continue with user studies to determine user tasks that are well-supported by our semantic search systems, and to identify gaps in data, services, and interface designs for meeting user needs. We may also prototype extensions to our platforms and metadata and test their effectiveness. Platforms we will use for evaluation include the SHARE-VDE linked data catalog being developed with various library partners, the Blacklight-based catalog developed at Penn for our local library collection and for an experimental Ivy+ shared discovery service, and Penn's Online Books Page catalog and accompanying Forward to Libraries service. We hope in our research to articulate the best uses of semantic searching, and the most effective investments in systems, interfaces, and metadata to support discovery in a linked-data environment.

Session recording

Metaphactory (https://metaphacts.com/) is a linked data software framework for managing, querying, and exploring knowledge graphs. It supports rapid prototyping and provides an environment for testing and applying search and discovery strategies driven by linked data. Among its features, metaphactory includes a built-in templating engine that facilitates the creation of custom search forms and visualizations. Yale University Library used funds from its Linked Data for Production (LD4P2) cohort subgrant to purchase a metaphactory license, and staff are currently in the process of receiving training from the vendor (metaphacts GmbH) in order to develop expertise in using the features of the software. As the only LD4P2 institution to use its subgrant funds to purchase a license for enterprise-grade software built entirely on linked data standards, Yale's experience with metaphactory is unique in the LD4P2 cohort. A demonstration and critical evaluation of the tool will provide attendees with a concrete example of the possibilities and potential limitations of using a discovery platform that is linked data native.

Session recording

Will Linked Data Enhance the Discovery of Multilingual Resources? Charlene Chou and Alexandra Provo (New York University)
Slides
Multilingual resources have been challenging for bibliographic description and discovery systems for years because of complex issues concerning data models and interface functions in various systems and platforms. For example, the IFLA Library Reference Model indicates that "in the implicit conceptualization of librarians, all translations are viewed as expressions of the same work; rights societies have a very different concept of "work", and regard each translation as a distinct "work"." For countries with multiple official languages, many publications are multilingual and distinctions between the original language and translations could be confusing for users. When the resources are written in both Roman and non-Roman scripts, bibliographic descriptions could get more intricate due to the display of native scripts, transliteration and translations. This presentation will explore and examine whether current metadata models, linked data editors as well as discovery systems would enhance the discovery of multilingual resources. Various use cases will be used to describe the challenges and opportunities for improvements. For instance, what are the issues of describing these resources in the Library of Congress BIBFRAME Editor and Sinopia Editor currently? Shall we adjust our approaches to accommodate these specific demands for better discovery? In addition, some use cases will be used to illustrate complex issues of digital collections at New York University, such as Arabic Collections Online (ACO) and the Hemispheric Institute Digital Video Library (HIDVL). In the case of ACO, the multilingual metadata needs of the digital library website are sometimes at odds with rules governing MARC-based cataloging practices. This case hints at future challenges we may face as we move legacy metadata to linked data environments, imposing new models and expectations on our metadata. As a tri-lingual project, HIDVL presents related opportunities to think through how we might improve storage of multilingual metadata. We will conclude with ideas for further research as well as recommendations and suggestions using the models of Wikipedia/Wikidata for multiple language interfaces.

GaNCH: Using Linked Open Data for Georgia's Natural, Cultural, and Historic Organizations' Disaster Response, Cliff Landis and Allyson F. Smith (Atlanta University Center Robert W. Woodruff Library)
Slides
Through a one-year LYRASIS Catalyst Grant, the Atlanta University Center Robert W. Woodruff Library is creating a publicly editable directory of Georgia's Natural, Cultural and Historical Organizations (NCHs), allowing for quick retrieval of coordinate location and contact information for disaster response. Directory information is being compiled, updated, and uploaded to Wikidata, the linked open data database from the Wikimedia Foundation. Directory information will then be delivered via a website, allowing federal, state, and regional emergency responders to quickly search for NCHs in disaster areas. Partnerships with state professional NCH organizations ensure that the compiled datasets will be sustained over time. Session attendees will learn how the directory was created, and how to replicate the project to preserve their own region's rich natural, cultural, and historical resources using free online tools.

Faculty Curation and Linked Open Data: BigDIVA.org, Laura Mandell (Texas A&M University Advanced Research Consortium)
Slides
The intended audience for this session are librarians and digital humanists interested in visualization tools that are being developed for Linked Open Data, especially those that enable serendipity in research and discovery of open-access materials via searches that integrate proprietary resources as well. The Advanced Research Consortium (ARC) is the overarching organization for NINES, 18thConnect, Medieval Electronic Scholarly Alliance (Mesa), ModNets, and three forthcoming data aggregators: the American Antiquarian Society's Networked Early American Resources, Disability Studies, and Asian American Literature. Each aggregator and scholarly community provides peer review for digital projects and approves ingesting other open-access materials provided by libraries and museums. Once accepted by the scholarly community -- nines.org, 18thConnect.org, mesa-medieval.org, modnets.org -- RDF metadata for each project, vendor, and institution is ingested into the ARC database, along with proprietary materials essential to researchers in those fields from JSTOR, ProjectMuse, Gale, and ProQuest, among others. The complete ARC catalog is visible through BigDIVA.org, the Big Data Infrastructure Visualization Application. ARC has partnered on a grant received from the Canadian Foundation for Innovation to transform the current ARC metadata (XML-RDF) into Linked Open Data and to transform Big Diva into an LOD visualization tool. We would like to present and receive feedback about our metadata, search capacities, the BigDIVA tool and our ideas for transforming BigDIVA into an LOD Viewer.