ORD-Xplore

ORD-Xplore aims to support digital librarians in the unification process of highly heterogenous digital editions so they can be supported in their goals for enhancing FAIR principles, leading to more open, connected, and accessible editions.

Digital editions are collections of historical document sources in digital representation, often used in digital libraries and research, e.g., in the digital humanities. Digital editions may form isolated islands of valuable open research data. However, due to digital editions’ domain diversity, heterogeneity, and lack of standardization, they currently face significant interoperability challenges. This problem is critical, as it prevents efficient access to information, findability, and reuse of digital documents from multiple editions. Our goal is to support digital librarians in their process of unifying digital editions. We present ORD-Xplore, an interactive visual data analysis approach that bridges digital editions from the UZH Zentrum für Digitale Editionen (ZDE) through three shared Connection Types: textual, temporal, and spatial. These Connection Types are the result of a careful abstraction process, achieved in a collaborative design study with domain experts in digital libraries. The resulting ORD-Xplore interface allows digital librarians to explore and study the connectivity of six digital editions in detail through its multiple linked views. In our evaluations, we find indications for a high degree of usefulness for librarians, including overall effectiveness and task success. ORD-Xplore is an important step towards wider accessibility of heterogeneous open research data in digital editions, promoting standardization and cross-edition interoperability.

First, digital editions are constantly growing, due to the contributions of ongoing digitalization and curation processes. Second, digital editions are heterogeneous, according to their domain, their context of use, and their document types. Third, digital editions have complex metadata standards. Using the studied TEI-XML standard as an example, librarians are tasked to make sense of 585 tags, many of which are ambiguously used. Fourth, unifying editions requires different granularities of support, from a single-document level to an edition level (collection of documents) to a collection of editions level (collections of collections of documents). Finally, different digital editions do not naturally offer semantically meaningful types of connections, and librarians quickly lose the overview of unifying factors due to edition diversity and complexity.

With ORD-Xplore, we aimed for the design and development of a web-based interactive visual exploration tool to allow digital edition experts to gain insights and guidance into the interoperability and unifying characteristics of these non-standardized digital editions.

The results of this project is the interactive visual ORD-Xplore interface consisting of five views that enable librarians conduct their tasks, addressing main challenges. The evaluation conducted so far indicates the high efficacy of ORD-Xplore to find commonalities between non-standardized, heterogenous digital editions, across all user groups consisting of digital librarians and non-experts in digital libraries on the basis of three connection types; time, textual structure and geo-spatial locations.

ORD-Xplore takes the high-level perspective to multiple digital editions, enabling digital librarians to compare editions by multiple promising types of connectivity: time, geo, and text. With the interactive visual ORD-Xplore interface, experts have a top-down perspective on the use of the TEI-XML metadata standard by diverse editions, revealing commonalities and differences, structural similarities, but also outliers and anomalies. Overall, ORD-Xplore can help guide digital libraries to find connections across digital editions and to reflect on the interoperability between digital editions.