Innovations in Information Retrieval: Perspectives for Theory and Practice

Citation

Lehane, R. (2012), "Innovations in Information Retrieval: Perspectives for Theory and Practice", Records Management Journal, Vol. 22 No. 2, pp. 140-141. https://doi.org/10.1108/09565691211268234

Innovations in Information Retrieval provides a window on contemporary developments in the world of information retrieval. Aimed at master's level students, it is a collection of essays by international academics discussing topical issues such as browsing, classification, folksonomies, and semantic navigation.

Two key themes that emerge in this book are that information retrieval is a faceted enterprise (different types of content demand different strategies for retrieval) and that it isn't just about producing the right result, the process of discovery itself is valuable.

That latter theme is explored most fully in a chapter by Richard Kopak, Luanne Freund and Heather L. O'Brien. These authors argue that information retrieval should be conceived of as an experience rather than simply as a transaction, because, in the process of trying to find things, users are learning, connecting, and creating. They describe this more holistic view of the information retrieval process as “semantic navigation”. They suggest that information retrieval systems be designed with this quality in mind. To this end, they make three general recommendations: show users the underlying model of the information system and its components (to allow users to make sense of those structures), provide tools for users to probe and to record their experiences (users are creators too), and take care with the design (because it isn't just about the result, it is about the experience too).

David Bawden is also interested in the capacity of information retrieval systems to do more than just retrieve things. He looks specifically at browsing interfaces. Browsing isn't just for bookshelves, Bawden argues, it remains a relevant strategy for online retrieval too. He reminds us that, while when we search we must search for something, when we browse we have the opportunity to make serendipitous discoveries. In this way browsing can foster creativity and innovation. Like Kopak, Freund and O'Brien, Bawden concludes with three recommendations for the design of information retrieval systems: show example contents of the system, reveal the system's structure, and highlight relationships between objects.

One of the most interesting chapters in the book is Charlie Inskip's introduction to music information retrieval. This is an area of research that has boomed over the last decade. With the widespread adoption of digital music players, consumers now have millions of songs at their disposal. How do they find the right music amidst this abundance? Music poses a set of unique information retrieval challenges: the target of searches is often unknown (find me something I will like), the nature of relevance is different, it can be tricky to identify “units of meaning”, the framing of queries is different (users often want to search by similarity), and music has unique metadata facets (pitch, tempo, harmony, timbre).

Anat Vernitski and Pauline Rafferty tackle another specialised area of information retrieval, fiction. They observe that it is hard to construct classification schemes for literature because works of fiction typically have many facets. They suggest a novel approach to retrieval based on references within works to other works (intertexuality). In the same way that scholars, particularly in the humanities, research by following citations from one source to another, users could jump between works of fiction based on their intertexual links.

Rounding out the book are chapters by Aida Slavic on classification, Isabella Peters on folksonomies, and Mike Thelwall on search engines.

So what's in this book for record keepers? It must be said that, apart from a short discussion of genre theory (which relates to diplomatics), not many of the examples referred to in the text are directly relevant. Nevertheless, retrieval is one of the key challenges for our field, an area where we are often failing users, and it is worthwhile looking to the information retrieval community for clues. Like music and fiction, records may have specific requirements but this does not mean that we need go it alone. There is much to draw inspiration from in Innovations in Information. For example, how can we use the intertextual links between records for discovery? Would searching by similarity be useful? What are the advantages of browseable interfaces for records? Above all, this book challenges us to consider retrieval systems as vehicles for more than just facilitating discovery. How can we build systems that foster creativity and learning, that allow engagement and interaction with records, and that expose the structures of record keeping systems? For considering questions like these, Innovations in Information Retrieval is a great starting point.