Making linked open dataeasy to understand & use
SHARE-VDE is a library platform that allows users to search in the data of a range of libraries at once. Technically the platform is innovative because it utilizes a new way of cataloging and linking data, called BIBFRAME. We are responsible for the user interface design of SHARE-VDE, and we focused on making the new way of storing data, useful and understandable for everyone.
About the project
The SHARE-VDE project is a collaboration between the Italian companies Casalini Libri and AtCult, and 20+ libraries in North America and Europe. Organizations involved in the project include Library of Congress, British Library, NLM, National Library of Norway, National Library of Finland, NYU, Stanford, UChicago, UPenn, Yale, Cornell, The Frick, Harvard, Northwestern, Princeton and more. Since 2018, we have worked with the community to analyze, redesign and rebuild the existing platform, with a focus on bringing it from a demonstration of technical functionality to a finished, user friendly product and a great user experience.
Library search engines today are almost exclusively searching in raw text data — data that does not contain links between e.g. the records describing conceptual works (like Franz Kafka’s “The Metamorphosis”) and versions of a work (like each of the 15 versions of “The Metamorphosis” that the library has in their collection). This absence of links between the various library records poses a number of challenges to how smooth of an experience can be offered to the end users (students, researchers). For one, it means that when you search for “The Metamorphosis”, the search engine is not necessarily aware that the 15 different copies — each published in different years, having different illustrations, translation and printing variations — are one and the same conceptual work. This results in the search screen presenting you with all 15 copies of the work, along with any other materials that might mention the word “metamorphosis” in their title, description or subject terms — whether or not the results actually have anything to do with the original work by Franz Kafka that you are looking for. If you happen to be interested in finding all copies of a work across multiple libraries, and not just in one particular library, the challenge only gets bigger.
This is where SHARE-VDE comes into the picture. SHARE-VDE is a search engine for searching in records that exist in a linked data format (also known as “Linked Open Data”). The data structure is based on a way of structuring data that is called BIBFRAME, but with some additional bells and whistles added on top to be able to provide an even better experience for people exploring the data. Since this great technical innovation will not make a difference to end users unless they can understand and easily navigate the interface for it, we worked with the Share-VDE community to 1) Make the BIBFRAME structure easy to understand for non-librarians, and 2) Make the BIBFRAME data useful to non-librarians.
To get an overall understanding of the project and the related concepts, we worked with our client to understand the product, reviewed existing documents related to the project as well as articles and websites dedicated to the concept of BIBFRAME and linked data.
After getting an understanding of the basics in the area, the existing product, and the desired features, we sketched out a rough suggestion for the interface of the product, which we then used to discuss with subject matter experts in the project. The SHARE-VDE project is structured with various groups contributing to the community around the project. We interviewed experts from some of these groups to learn about their view of linked data, its opportunities and challenges, and to gain insights into how we might make the system function optimally for end users. We had recurring meetings with subject matter experts to discuss the features, flow and layout of the interfaces, tweaking the design based on our discussions with the library staff members in the groups.
The research and collaborative design process led to the creation of a rough prototype of the system as we imagined it in the long term. We used this rough, long-term prototype to discuss the priorities, technical implications and feasibility of the various features for the first version of the product with the group of SMEs as well as the developers in charge of turning the design into a real, functional system. Based on those discussions, we worked out a short-term version of the same design, removing some features and keeping others based on what was easier to make and what we assessed was more important for the students and researchers who would be using the system.
With a rough prototype of the first version of the product, we moved on to create a refined version of the design for the developers to use in developing the interface and functionality. This design was again reviewed with the group of SMEs, which revealed further details to tweak.
With a detailed prototype, our team developed a set of interface components into a SHARE-VDE framework, which could then be put together to create the various relevant screens for the product by the development teams involved.