Presenting bibliographic families: designing an FRBR-based prototype using information visualization
| Published date | 09 May 2016 |
| Date | 09 May 2016 |
| DOI | https://doi.org/10.1108/JD-01-2015-0001 |
| Author | Tanja Merčun,Maja Žumer,Trond Aalberg |
| Subject Matter | Library & information science,Records management & preservation,Document management |
Presenting bibliographic families: designing an FRBR-based prototype
using information visualization
Introduction
Online bibliographic information systems have brought users the possibility to search and explore library
collections anywhere and anytime, but they have also been repeatedly characterized as difficult to use, frustrating,
inefficient, and missing the required functionality. Losing the position of primary information providers due to
their inadequate services in the changing information environment (Coyle and Hillmann, 2007), information
systems such as library catalogues would need to better support users in their information seeking process and
bring out the richness and value of bibliographic data in order to regain their value. While the summons for better
bibliographic information systems have been made in literature throughout the last three decades (Hildreth, 1985;
1989; Borgman, 1986; 1996; Bates, 1986; 2003 etc.), changes have been slow and have actually come to a hold in
the 1990s. In the last few years, competing services on the web and demanding web-savvy users have forced the
library community to reconsider its services, which has led to some noticeable modernizations in the way
bibliographic information systems provide search, presentation, and navigation. Despite the changes, however,
current linear displays of records still do not fulfil all the objectives and do not have the adequate structure for
representing as well as making navigable the “multidimensional complexity of the bibliographic universe”
(Bianchini, 2010). Approaching a turning point, some believe that catalogues will need to “change or perish” as
the current state will only remove them further from readers and collections (Bianchini and Guerrini, 2015).
Our research therefore sought an alternative design to current bibliographic information systems that would
not only improve the presentation of and navigation within bibliographic records, but also structure displays in a
way that would facilitate user tasks and understanding of the data. Exploring possible approaches to address the
identified issues, the following concepts have been chosen as the building blocks of our work:
a) FRBR (Functional Requirements for Bibliographic Records). Published by the International Federation of
Library Associations (IFLA, 1998), FRBR is a conceptual model of the bibliographic universe that was
developed to improve the cataloguing process as well as online library catalogues (Carlyle, 2006). It holds
the potential for creating more efficient and user-friendly bibliographic information systems by bringing
together all editions of a work that exist in a collection as well as by creating relationships between
various entities, such as related works or editions. While some attempts have been made on implementing
the model (overviews in Yee, 2005; Zhang and Salaba, 2009; Pisanski and Žumer, 2007), the experiments
so far have used only a fraction of FRBR potential. This means that even FRBR-inspired catalogues have
not provided a suitable structure to identify or navigate the majority of relationships that exist between
bibliographic resources. Our research used the entity-relationship model proposed by FRBR to restructure
existing bibliographic data. It also looked at how the converted data could be best presented in a user
interface to create an informative overview and support semantic exploration.
b) Work family. Based on FRBR and other literature on organizing and presenting bibliographic data
(Fattahi,1996; 2010; Ercegovac, 2006; Carlyle, 1997a), the concept of a work family was set as the main
element of our research. We have therefore not looked at the complete user interface design, but only at
how the entities and relationships within a work family could be brought together in a single display,
collocating different editions and versions of the work, showing related work, and providing a broader
exploration of works by or about an author.
c) Information visualization. Traditional linear lists of records do not provide the necessary structure that
would naturally support meaningful clustering or exploration of relationships (Carlyle, 1997a). To
overcome these limitations and provide a frame that would better suit the proposed model of a work
family presentation, a dynamic graphical display was chosen. As there are various information
visualization techniques available, each having its pros and cons, four alternative visualization techniques
have been selected in order to test which of them would be most appropriate for our model.
Combining the three concepts, the first objective of our research sought to propose an interface design for
presenting FRBR-based bibliographic data using work family grouping and hierarchical visualization. While the
proposed design was later also implemented and tested with users, this paper reports on the first segment of the
research that looked into information architecture and investigated possible display options. Deriving from the
objectives of bibliographic information systems, the introductory part analyses past and current literature to
identify their unfulfilled potential and possible solutions. The second part of the paper is dedicated to FRBR and
various questions related to its implementation, while the third part describes the design of a prototype system
FrbrVis. It explains the underlying ideas behind the prototype system and shows an example of how work families
could be presented and interacted with using information visualization.
Theoretical background
Problematic presentation and navigation of results
Online library catalogues as well as other bibliographic information systems have often been characterized
as frustrating, unintuitive, time consuming, and difficult to use and find relevant information. Literature on the
topic (e.g. Borgman, 1996; Bates, 2003; Novotny, 2004; Yu and Young, 2004; Rethinking…, 2005) describes
these problems and looks for reasons behind them. While not the sole factor, display and navigation seem to be
among the biggest problems in search systems today.
Carlyle (1999) emphasizes that poorly designed displays contribute to search failure and, in one of her
works (1997b), concludes: “… the disorganization of current online catalogue displays may cause users to
abandon their searches, leaving the catalogue frustrated and confused. Even experienced catalogue users may be
thwarted in their searches for particular authors and works by the lack of organization that exemplifies current
online catalogue displays.” She suggests that online catalogue displays are inadequate for searches where many
items are retrieved and proposes the use of summary displays that would classify or cluster search results and thus
enhance users’ ability to identify items of interest. She also points out that “displays that do not demonstrate
relationships among related items retrieved may leave users, at best, disgruntled over the amount of time
necessary to find what they are looking for and, at worst, oblivious to the fact that the library actually holds the
very item or items they seek” (Carlyle, 1996). Already in 1997 she concludes that changes are long overdue and
summons the library community to provide relationship-based displays that could be created using various
approaches such as graphical, hierarchical tree-structures (Carlyle, 1997a).
To continue reading
Request your trial