Introduction

Services in Web 2.0 generate a large quantity of information, distributed over a range of resources (e.g. photos, URLs, videos) and integrated into different platforms (e.g. social bookmarking systems, sharing platforms (Peters, 2009). To adequately use this mass of information and to extract it from the platforms, users must be equipped with suitable tools and knowledge. After all, the best information is useless if users cannot find it: ‘The model of information consumption relies on the information being found’ (Vander Wal, 2004). In Web 2.0, the retrieval component has been established through so-called folksonomies (Vander Wal, 2005a), which are considered as several combinations of an information resource, one or more freely chosen keywords (‘tags’) and a user. Web 2.0 services that use folksonomies as an indexing and retrieval tool are defined as ‘collaborative information services’ because they allow for the collaborative creation of a public database that is accessible to all users (registered, where necessary) via the tags of the folksonomy (Ding et al., 2009; Heymann, Paepcke and Garcia-Molina, 2010).

Folksonomies in information retrieval serve mainly to provide access to information resources for all users – where the initial motivation for tagging is often of a personal nature:

Metadata in the form of tags, titles or descriptions are especially invaluable for the retrieval of non-textual information resources such as images or videos because the state of the art of search engine technology allows only for textual research, meaning that these resources would otherwise be irretrievable (Dye, 2006). In these instances, it is also important for the retrievability of resources to be guaranteed by a broad-access vocabulary, and this is where folksonomy can provide excellent service (Goodrum and Spink, 2001, 296). Because of their strong user-orientedness and broad-access vocabulary, folksonomies can also facilitate the retrieval of relevant information in company intranets or in ‘enterprise search’ (Schillerwein, 2008).

Introduction

This chapter considers fiction retrieval research and initiatives, providing an overview of some of the approaches that have been developed. In particular, it describes two recent approaches to fiction retrieval that have made use of theoretical concepts drawn from literary theory.

Fiction is an interesting information domain because it includes documents that serve two purposes, which are reading for pleasure and scholarly study (Beghtol, 1994, 3), but fiction retrieval has not always focused on both aspects. In the 19th century, the approach was to treat fiction from a knowledge perspective within general classification schemes. The Dewey Decimal Classification (DDC), the Library of Congress Classification (LCC) and the Universal Decimal Classification (UDC) contain classes for literature, with the main subdivision in each case being the language in which it is written. Further subdivision is possible based on literary form, historical period or the works of an individual author (Riesthuis, 1997).

It may be that there is a relationship between the development of literary studies as academic scholarship and approaches to fiction retrieval in libraries. In the UK, for example, Literary Studies as an academic discipline developed in the 19th century, institutionalized in the Mechanics’ Institutes, workingmen's colleges and extension lecturing circuits before its acceptance in the universities (Eagleton, 1996, 23). In its early days in the academy, English Studies was considered to be ‘an upstart, amateurish affair as academic subjects went, hardly able to compete on equal terms with the rigours of Greats or philology’ (Eagleton, 1996, 25). It is a relatively new academic discipline, although its research methods and interpretative approaches have a much longer history, at least since Aristotle's Poetics. We know that during the late 19th and early 20th centuries public libraries in England were still capping the number of novels that patrons were able to borrow, which suggests some nervousness about the role of the library and the librarian. There seemed to be concern, evidenced in the letters pages of the Library Association journal of the time, about whether libraries should be encouraging the reading of novels and the consumption of fiction for pleasure alone or whether they should be concerned with the intellectual ‘improvement’ of their patrons by encouraging the reading of serious, knowledge-based literature (Barr and Sturges, 1992).

Introduction

In this chapter we focus on the research area of digital information interaction, which emphasizes searchers’ direct engagement with and manipulation of information objects as they search and browse through digital information environments. This is an area of active research that has opened up in recent years as information retrieval (IR) research has expanded its focus from the mechanics of retrieval (i.e. indexing, data structures and retrieval algorithms) to include a broader ‘retrieval in context’ perspective that takes into account the whole system, the affective, cognitive and physical attributes of users and the environment in which searching takes place (Ingwersen and Jarvelin, 2005). A number of meetings and workshops have focused on this area, including the Information Retrieval in Context (IRiX) workshops at the ACM SIGIR (Association for Computing Machinery Special Interest Group Information Retrieval) conference (2004–5), the Information Interaction in Context (IIiX) Conference (2006–ongoing) and the Human Computer Information Retrieval (HCIR) Workshops (2007–ongoing).

Immersive information systems, in which users are placed in content-rich environments with tools and technologies designed to help them search, filter and navigate through the information landscape, are now commonplace. In these types of environments, IR can no longer best be modelled as a transaction, but is more akin to an ‘experience’, in which searchers are influenced by the information objects they encounter and also shape and create their information environment by selecting, linking, tagging and commenting (Marchionini, 2008; Toms, 2002). For many people, these kinds of interactions with digital information are the primary means by which they read and learn, whether at work, at school or at leisure.

However, numerous studies that utilize server logs of bibliographic and web based information systems have shown that digital information behaviour is markedly different from that in traditional print information environments. Digital information use tends to be non-linear, shallow and piecemeal. Readers jump frequently from one document to the next, ‘skimming and bopping’ through networked information spaces (JISC, 2008; Nicholas et al., 2007). Concerns have been raised that this type of information behaviour represents the ‘intellectual equivalent of empty calories’ (Rich, 2008) and inhibits deep, critical engagement with informational content (Marshall, 2005).

Introduction

Information Retrieval (IR) research typically evaluates search systems in terms of the standard precision, recall and F-measures to weight the relative importance of precision and recall (e.g. van Rijsbergen, 1979). All of these assess the extent to which the system returns good matches for a query. In contrast, webometric measures are designed specifically for web search engines and are designed to monitor changes in results over time and various aspects of the internal logic of the way in which search engine select the results to be returned. This chapter introduces a range of webometric measurements and illustrates them with case studies of Google, Bing and Yahoo! This is a very fertile area for simple and complex new investigations into search engine results.

The modern commercial web search engine is a highly complex system (Arasu et al., 2001) with vast social and commercial significance (Van Couvering, 2004, 2007). Although they can be evaluated to some extent with traditional IR measures like precision and recall, web search engines behave differently from traditional IR systems in many respects. Three key differences are the importance of rank order in the results; the limitation to 1,000 matches per query; and the goal of delivering relevant and useful results rather than technically accurate matches (in the sense of: Bar-Ilan and Peritz, 2008). In response, evaluation metrics have been developed to measure new characteristics of web search engines, such as the effectiveness of the rank order of the results (Zaragoza, Cambazoglu and Baeza-Yates, 2010), mean average precision (Turpin and Scholer, 2006) and discounted cumulative gain (Jarvelin and Kekalainen, 2002). A broad common goal is to assess the extent to which any web search engine delivers good and relevant results to users. For some information scientists, however, IR goals are not sufficient with regard to web search engines, for two reasons.

First, search engines like Google are so important in academia and daily life that it is important for information professionals to understand something of how they work and what their limitations are.

Introduction

The vision of the semantic web is gradually unfolding and taking shape through a web of linked data, a part of which is built by capturing semantics stored in existing knowledge organization systems (KOS), subject metadata and resource metadata. The content of vast bibliographic collections is currently categorized by some widely used bibliographic classification and we may soon see them being mined for information and linked in a meaningful way across the web. Bibliographic classifications are designed for knowledge mediation, which offers both a rich terminology and different ways in which concepts can be categorized and related to each other in the universe of knowledge. From 1990 to 2010 they have been used in various resource discovery services on the web, and they continue to be used to support information integration in a number of international digital library projects. In this chapter we will revisit some of the ways in which universal classifications, as language-independent concept schemes, can assist humans and computers in structuring and presenting information and formulating queries. Most importantly, we will highlight issues important to understanding bibliographic classifications, identifying both their unused potential and their technical limitations.

Background

Classifications are created using the intellectual power of ontological observation and analysis and reflect whichever approach one decides to take in defining and grouping things. Whatever the motive for their creation, classifications are logically and semantically organized schemes of concepts representing some kind of reality and, as such, are essential and versatile tools for the representation and visualization of a knowledge space. A classification, by nature, cannot serve all purposes; however, if it is universal with respect to knowledge coverage and structured in a comprehensive, detailed, highly elaborate and logical way, it is likely to be more versatile than a scheme created for one single subject, system or task.

Bibliographic or documentary classifications are a special kind of knowledge classification designed for the mediation of knowledge. Their unique feature is that they are not concerned with objects or entities, as other knowledge classifications, but rather with subjects, i.e. the ways in which entities are described in documents. Bibliographic classifications will systematize phenomena and the topics that can be studied in relation to these phenomena, but will also provide the vocabulary necessary to denote types of knowledge presentation, points of view, the targeted audience or form of document.

Introduction

There is a long history of music librarianship in the domain of printed Western classical music. Special schemes have been developed to aid in the organization and retrieval of musical works, and existing schemes have been widely used to include these types of documents in larger physical library collections. However, the advent of digital consumer technology in the form of MP3 players and mobile phones, combined with the enormous impact of the internet and the digitization and ease of compression of audio files, has brought new formats and types of user interaction to the fore. This has led to an explosion in music information-retrieval research, concentrating on how most beneficially to use computers to organize, search and retrieve music information and recordings from large digital collections.

Many of us today carry around music collections of thousands of digitized music recordings and access all manner of types of music on the web, but still are unsure what to listen to next: the enormous size of these collections and the instant accessibility of 8 million Western pop, classical, jazz and folk songs can cause confusion and trepidation. Where the classical music researcher would previously have consulted academic texts and visited a specialist music library, or the post-rock listener would have read the New Musical Express and visited the Rough Trade shop for advice on what was coming up, now we access music through hand-held devices and laptops. The issue is no longer ‘I hope I can find that Velvet Underground live album somewhere this year, I wonder what it sounds like’, but ‘Which Velvet Underground live track shall I read about/ download/stream now?’

Efficient and effective automated retrieval tools can help users to access globally distributed and personal digital music libraries. Recent research in this area generally involves extracting musical features from audio files and using these features to make comparisons between pieces of music in order to determine whether or not they are similar. Enormous data resources from internet usage and unfeasibly large music collections are used for machine learning to enable auto-tagging. The web is trawled for socio-cultural information, and algorithms are developed and tested by a thriving scientific community.

Change for librarians is not new. New information sources and formats have regularly been assimilated into library collections as librarians endeavour to meet their patrons’ information needs. Microfiche, microfilm, CD-ROMs, web pages and social media content have all been introduced into library collections over the years. Increasingly however, as well as more varieties of documents, there are large quantities of data being made available online in a range of formats that require new skills and knowledge on the part of librarians. Over the past couple of years governments around the world have started making vast quantities of non-personal public data available online, research-funding agencies are placing greater emphasis on researchers making data available in digital repositories, industry is making various data sets available in the hope of making use of external talent and an increasing proportion of the public are marking up information on their own websites in a standardized way so that it can be read automatically. Together this data combines to make a ‘web of data’ – structured data in a machine-readable format that is being made publicly available online by individuals and organizations from every sector of society. This data potentially has huge economic, research and social value, and library and information pro - fessionals across the different sectors have a pivotal role in making sure that this value is realized.

The aim of this book is to provide an introduction to the web of data for the library and information professional, emphasizing: the importance of open data; the broad spectrum of individuals and organizations publishing the data; the meaning of semantic data and its advantages; the various publishing formats that are being used; and most importantly, the role of library and information professionals as facilitators of access to this web of data. This book is not designed as a how-to guide for the publishing of data online or for including semantic markup within HTML (HyperText Markup Language) pages, although some of the examples will help.

Introduction

In his Open Secrets article for The New Yorker, Malcom Gladwell (2007) reiterates the US national security expert, Gregory Treverton's distinction between a puzzle and a mystery. With a puzzle, additional information is required for it to be solved, whereas with a mystery, part of the problem is too much information. The primary focus of Gladwell's article is the Enron scandal of 2001, where Enron's assets were found to have been greatly inflated. Much of the information was in the public domain, although with so much information available, the disparate parts had not been pieced together. The vast quantity of information now available is widely recognized. As Eric Schmidt stated in August 2004: ‘There was 5 exabytes of information created between the dawn of civilization through 2003, but that much information is now created every two days and the pace is increasing’ (Kirkpatrick, 2010). It was also widely reported in 2010 that the world would pass an estimated 1.2 zettabytes of information, the equivalent of 75 billion iPads (Blake, 2010). Such numbers are barely comprehensible, although what is clear is that if this information is to be used to its full potential it is important that we find ways of reducing the need for human input in bringing together disparate information in useful ways.

This chapter looks at the methods that have evolved to deal with the increasing amount of information available online at both the document and the data level. Over the last two decades we have moved first from directories to search engines, and then more recently to the crowd-sourced solutions of tagging and tapping into the streams of our online ‘friends’. Such methods have helped to deal with much of the unstructured content available today, although there is a significant gap between what we can achieve online today and what could be achieved if the meaning of more of this information could be understood without the need for human interpretation.