Abstract

This paper reports on an ongoing doctoral research project that aims to explore the retrieval of digital art in a Digital Library (DL) context, by assessing the information needs of Arts & Humanities scholars. Using Web Service technology, the findings will be instantiated in DL tools. In this paper, it is argued that the products of digital art are a source of learning and research in the disciplines of Arts & Humanities, provided they are systematically made accessible for long-term retrieval. The paper presents the research background and the objectives of the project, then sets off to demonstrate the user needs assessment methodology. The final section exhibits the contribution of the project to the broader Digital Library research field and offers pointers to further discussion on the subject.

Keywords

digital art, information retrieval, representation, user needs assessment, web services

1 Introduction

Digital Libraries (DLs) as resources of scholarly material have stimulated the interest of researchers from a broad range of disciplines. From the technical, practical and managerial issues pertaining to the digital library, to the notion of the digital library as a portal to digital/digitized collections, this vivid activity is underpinned by a growing demand from user communities for services and content that extend beyond the interim state of accessing the digital equivalent of a physical library catalogue.

This paper presents a doctoral research project that explores the retrieval of digital art in the context of a digital library infrastructure. The motivation for the project lies in four issues concerning digital art and DL research. The first is the importance of user-oriented studies and user-driven DL-service development [4, 12, 14, 18, 23]. The second is the service-oriented architecture (SOA) of Digital Libraries, which allows DL applications to consist of self-contained services so that each provides different functionality [7]. The third is that a limited number of institutions have engaged in efforts to systematically collect new and uniquely digital artefacts as part of their informational material [24]. A current challenge is the requisition for long-term access to a diverse array of digital content that can act as evidence of contemporary culture and social change brought by emerging technologies for future scholars [22]. The fourth issue is that at present, the potential value of digital art is undercut by the lack of services that aggregate the dispersed material of on-line resources dedicated to such objects into a user-defined dataset, consisting of components that can be searched, retrieved, compared and associated with each other. The aforementioned issues define the research problem explored in this thesis, namely:

Based on the needs and expectations of the digital art user communities, what type of services to provide for Digital Libraries by employing a service-oriented and loosely-coupled architecture.

Building on the problem statement, the presented project is further guided by the hypothesis that:

The products of digital art can form a cogent source of learning and research in the disciplines of Arts & Humanities, so long as they are systematically collected and maintained in organised electronic repositories.

Whereas research in other forms of digitally-born art has been rigorous (e.g., 3, 17, 20]), the digital art domain remains largely unexplored. It is therefore crucial to make such material readily available and attend to the particular characteristics of a new and unestablished medium of artistic expression that is becoming increasingly popular with artists and scholars alike.

Based on the findings of a survey involving Arts & Humanities scholars, the goal is to design and develop services that adhere to the expectations of the identified user community for accessing digital art artefacts. If Digital Libraries are envisioned as a distributed information environment, a service-oriented architecture is the key in achieving the indicated functionality. To this end, the project employs the Web Service architecture as a web-enabled technology for communication within distributed systems that offers distinct benefits. The objective is to evaluate an open solution for representing, managing and using digital art in digital libraries in a distributed context.

The purpose of this paper is to provide an overview of the current progress of the project and exhibit the methodology used to address the problem. Section 2 reviews the research background and briefly explains three concepts regarding the research problem that set the groundwork for the project. Section 3 presents the methodology adopted and analyses the particulars of the survey for the user needs assessment. The final section summarises progress to date, exhibits the contribution of the thesis to the broader field of Digital Library research and offers pointers to further discussion on the subject.

2 Research Background

In recent years, the field of Digital Libraries has witnessed a growing trend to assess the efficiency of the offered services, as perceived by the creators, field specialists, targeted audiences and funding bodies. Front-end evaluation in the form of a user needs assessment has been employed in a number of projects, yielding interesting results (e.g., [4, 14, 15, 19]). Needs assessment describes "a systematic and on-going process of providing usable and useful information about the needs of the target population – to those who can utilize it to make judgments about policy and programs. [It] is population-specific, but systematically focused, empirically based, and outcome oriented" [21]. To avoid ambiguity with similar concepts, it is accepted here that a need represents a set of essential elements to meet an objective; each element might be either a "condition or capability needed by a user to solve a problem or achieve an objective" [10] (user requirement) or a "defined criterion used as the basis for information systems design" [10] (system requirement).

A challenge in the front-end evaluation processes is that the subjects provide feedback based on their speculations about a set of services that are yet to be developed; the success of the endeavour is largely dependent on the context in which the findings will be interpreted. In this research project, the context of the user needs assessment was determined by defining (1) the Digital Library as a sociotechnical network; (2) Arts and Humanities scholars as the survey informants; and (3) digital art as the information object. The following sections provide a short analysis of how these research areas influenced the research context and set the background for designing the user needs assessment phase.

2.1 Sociotechnical Networks

The interdisciplinary nature of Digital Libraries poses an explicit difficulty in studying this technology, in that information is scattered among the writings of various thematic categories, so much so that reaching for a definition largely depends on the discipline through which it is expressed [11]. In this sense, a single definition cannot capture the complex notion and diverse aspects that correspond to the DL phenomenon [5].

Watson-Verran and Turnbull [25] argue that Digital Libraries are "the amalgam of places, bodies, voices, skills, practices, technical devices, theories, social strategies, and collective work that together constitute technoscientific knowledge/practices". Hence, Digital Libraries encapsulate the combined work of three constituents: technical infrastructure, collections of digital objects and people. The theory of Social Informatics epitomises this view and considers the Digital Library as a sociotechnical network that connects equipment, software and hardware vendors, technical specialists, managers, ICT policies and funding with the people who use it in their everyday life [11]. If DLs are to be perceived as socially shaped technologies, the aggregation of digital objects in their collections rests upon socially generated incentives such as education, knowledge, accessibility, preservation of collective memory and sharing of cultures and ideas. The notion of the sociotechnical network embraces the orientation of user communities that are represented by a mutual anticipation for sharing and retrieving scholarly material with an interactive quality between contents and contexts.

With respect to the research problem, the study of Social Informatics revealed fine-grained details. The formulation of contextual research questions that move beyond simple determinism significantly aided in giving a more tangible form to the original problem statement, while the idea of the sociotechnical network emphasised the role of the end-users in the design of services. More importantly, Social Informatics highlighted a factor that forms the cornerstone of the presented project: that any application is designed and implemented to be ultimately used by people and support their needs.

2.2 The Survey Informants

In the Digital Library environment, the purpose of accommodating learning and research, as defined in [7], is the linchpin that unifies a community of new and heterogeneous sociology [2] that nurtures a common ground of beliefs, ideas, behaviours and informational needs. In this sense, the selection of the informants that would participate in the user needs assessment process was founded on the level of their motivation to complete the survey and share their views. Aldridge and Levine [1] suggest a set of reasons why people are willing to take part in surveys. In brief, the relevance between the subject of the survey (in this case the user needs assessment) and the experiences, education and line of work of the participants influences the degree of their willingness to volunteer for the survey.

Thus, the decision to address Arts and Humanities scholars as the user needs assessment informants was based on four criteria: (1) the immediate interest in using digital art as part of their learning and research; (2) the clear benefit from services dedicated to performing functions on digital art objects; (3) the relatively common level of expertise in the subject area; and (4) a fair understanding of the research problem and the impact of their participation in the survey.

2.3 Digital Art

Digital Art can be defined as the generation of artistic expression in digital format, created by means of computers. This particular digital object genre is distinctively different from – although often confused with – digitised art. Digital Art is inherently digital, in that the primary creation of the image does not rest upon the use of traditional painting equipment; instead, computer graphics software combined with the talent of the artist deliver the final product. Digital art amalgamates traditional and contemporary techniques; Jarvis [9] describes this feature as a "chameleon-like ability" that enables the products of digital art to mimic traditional artistic styles. As a result, new subtypes of computer-generated artistic creation have emerged – digital painting, rephotography and software art, just a few to name.

Some isolated studies (e.g., [8]) have attempted to investigate the problematic areas regarding the description, cataloguing and--consequently--retrieval of digital art in organised repositories. The general conclusion is that extant metadata standards cannot straightforwardly accommodate the particular characteristics of this medium. This deprives digital art objects of the fundamental features that distinguish them from other art forms and has serious implications in the allocation of descriptors and the assignment of identifiers. A good example of such an occurrence is the case of digital paintings; lack of sufficient descriptive metadata can lead to them being confused with traditional artwork on canvas that digital painting so closely assimilates. This problem becomes more evident with a longitudinal view in mind, when traces of the original nature of the material are potentially lost.

At the same time, the case with which information is deployed and circulated in the online environment, alongside with the lack of a legal framework that covers copyright issues globally, nurtures a cognitive and social culture that is based on what Manovich [13] calls the basic logic of modern cultural production: "download images, code, shapes, scripts, etc.: modify them and then paste the new work online – send them into circulation". This creates an impetus to safeguard the originality of digital art material and cater to long-term preservation, especially in the case of a Digital Library where the contents have to be creditable. Building on these issues, the presented project argues that a distributed environment that brings together the holdings of various digital art repositories can promote the authentication and originality of the objects through uniformity of the same description in different catalogues and collections.

3 Methodology

The complexity of the research problem calls for a substantial and well-defined methodology that takes into account the individual characteristics and properties of each research objective and follows a progressive, top-down approach. To this end, this work is divided into three phases. The first phase--Literature Analysis--is dedicated to understanding the context of the research, as elucidated through the relevant literature, existing knowledge and previous research findings. It starts with the investigation of the theoretic background of social practice in evaluating digital libraries and designing services, so as to highlight underlying aspects pertaining the format of the user needs assessment and the selection of services to be eventually implemented. The goal of this step is to prove that social context does matter when designing and developing technologies, especially for use with material such as art, which is ipso facto socially shaped. The attention then turns to the particulars of current information retrieval practices in Digital Libraries. Although the research has a specific interest in digital art, appraising retrieval tools of other types of collection material in the form of a comparative review provides a guide for the user needs assessment and a measure for the respective findings. In parallel, the evaluation process of prominent Digital Library projects is reviewed and documented to contribute towards an enhancement of knowledge on the subject based on real-world examples.

The second phase of the research – User Needs Assessment – is devoted to the design of a strategy to elicit the needs and expectations of Arts and Humanities scholars for retrieving digital art that draws on the issues explored in the previous step. The length of the process is analogous to the sample size and the selected sampling method. The last phase -- Web Service Design -- builds on the findings of the user needs assessment and delves into the design and development of the service modules that provide the identified functionality for information retrieval. A summative evaluation process gauges the success of the final products and sets the ground for the proposal of further augmentations and research conclusions.

3.1 User Needs Assessment: Target Populations and Research Questions

The analysis of the research problem revealed three survey populations to participate in the user needs assessment, that could inform the research on (1) the retrieval needs of digital art; (2) the specific requirements of digital art objects in terms of description and cataloguing within a collection; and (3) the potential incentives that can lead to digital art inclusion in a Digital Library environment. Specifically, the survey addresses:

• Postgraduate students in the disciplines of Arts & Humanities
• Digital artists
• Representatives of digital library institutions

Given the lack of previous user-centred studies focusing on digital art, the research questions reflect the exploratory nature of the user needs assessment and were designed to provide insight in a wide range of considerations and initial speculations with respect to representation and retrieval. To this end, a separate list of questions was constructed for each target population, thus allowing for the elicitation of fine-grained details not commonly shared among the three groups, and helping focus the development of the data collection instruments.

Arts and Humanities Students
The issues of interest here centre on the educational outcome of digital art retrieval as perceived by this target population. What is interesting in this case is to elicit information that Arts & Humanities students are most knowledgeable of and falls within their expertise. Thus, the research questions for this group summarise into the following:

1. What is the level of awareness over the nature and content of digital art?
2. Has their learning/research involved the use of this medium? If so, in what context and with what degree of relevance to their field of study?
3. What is the degree of differentiation between digital and digitised art and in what manner can this information be represented?
4. What search patterns have been (or would be) adopted in order to retrieve digital art objects? If digital art has been retrieved in the past, to what extent do system representations of the objects match their information needs?
5. What is the level of awareness of online organised repositories as Digital Libraries? If a digital library has been used in the past, in what context did this occur and for what purpose?

Digital Artists
As the creators of the observed digital objects, digital artists can substantially communicate their experience in deploying their work, describing the material and safeguarding its originality. The research questions for this group include:

1. What are the preferred methods for publishing a work of digital art? Which repositories/online resources do they involve?
2. What are the particular characteristics that distinguish digital art from other forms of art? How do digital artists provide a representation of their work? Are existing metadata standards adequate to formally and effectively describe this digital object?
3. How is the originality and uniqueness of digital art artefacts reinforced? How is copyright secured?
4. Would a distributed environment that combined resources from different repositories in a single dataset be expedient? What are the prerequisites of such functionality?
5. Do digital artists perceive their work as expressed in digital objects and metadata fitting in the context of a digital library?

Digital Library Representatives
Although the focal point of the presented project are the needs of end-users, there would be little significance, if none at all, to examine the representation and retrieval of digital art, if Digital Library institutions were reluctant to approach such material as part of their collections. Hence, the research questions addressed to this particular group have been designed to provide supplementary details on the current state of affairs with respect to digital art in digital libraries and highlight potential future trends. In particular:

1. What is the current level of digital art incorporation in DL collections? What are the motivations/prohibitions behind this decision?
2. What are the prerequisites for investing in incorporation of digital art in DL collections?
3. Does digital art have particular representation and retrieval needs? If so, what are these needs and how can they be accommodated in the DL context?
4. Is the current level of services adequate to cater to digital art retrieval?
5. How can the issue of originality and reliability of digital art objects be resolved?

3.2 User Needs Assessment: Data Collection

In order to focus on the representation and retrieval needs of digital art as expressed by the target populations, a means of corroboration was necessary. To provide this information, individual data collection instruments were designed for each identified group of informants that resonate with the issues addressed in the respective sets of research questions.

In the case of scholars, the population was limited to postgraduate students in the disciplines of Arts and Humanities who study in higher education institutions situated in the United Kingdom. The population complies with the criteria for the selection of survey informants displayed in Section 2, and ensures a high level of coherency with respect to the social context in which these informants live and operate. The sampling frame was identified as Arts and Humanities postgraduate students in Scottish higher education institutions, as representative of the population. This is particularly important for the reliability and validity of the survey findings and allows for generalisations to be drawn to the broader survey population. Using a confidence level of 95% and a confidence interval of 4, the size of the sample was calculated to 404 participants¹.

The data collection instrument for this sample took the form of a computer assisted (online) questionnaire, which followed a semi-structured approach. Emphasis was given to qualitative analysis that allows for more nuanced information to be gathered. McCullough [16] suggests that online surveys are time- and cost-effective and can generate rather accurate (statistical) information. These benefits are extenuated by the potential problems of online data collection identified by Zhang [26], especially as far underrepresentation of certain groups is concerned. Although a standardised sampling method could not be applied in this case, the identified sample is representative and unbiased – as it can be safely assumed that all students in Scottish Universities have access to the Internet and the participants are randomly self-employed. The questionnaire was evaluated in two phases. During the first phase, the data collection instrument was examined by a panel of experts. In the second phase, the questionnaire was pilot tested by a sample of 20 participants. The results of the evaluation were incorporated in the updated and final version. The questionnaire was available for completion for a period four months, starting in March 2007.

To further complement the findings of the survey, twenty-five open interviews with digital artists were organised. A basic criterion for this selection was the experience of the artists with a strong emphasis on a teaching/academic background. As far as representatives of Digital Library institutions are concerned, five open interviews took place concurrently with the questionnaire and addressed IT specialists and digital librarians employed in DL institutions.

4 Contributions and Discussion

At the current stage of the research, the data collection and analysis processes have been completed, with the former having gathered the required number of responses. A case study of eight indicative DL evaluation efforts in extant projects that was carried out prior to survey design significantly contributed to the success of the data collection instruments by illustrating the role of strategic decision-making in terms of evaluation methodologies, selection of metrics and reporting of evaluation results².

The statistical interpretation of the study participants' feedback revealed recommendations for metadata and services currently not provided by digital art repositories. The data analysis provided an evaluation of existing searching, retrieval and material description provisions that act as a benchmark for the development of the web service modules. In addition, the survey results allowed for an appreciation of the level of awareness and need to employ digital art as scholarly material in academic learning and research that reinforce the validity of the original research hypothesis. The project will culminate with the development and evaluation of services offering the identified functionality and the correlation between the results of the front-end and summative stages that will test the validity of the hypothesis and signify the level of the project's successful completion.

The major contributions that are expected to derive from the presented project for the broader Digital Library field are open to discussion and extend an invitation to other researchers to join the effort for new media research in Digital Libraries. These contributions are summarised in the following points:

• The project will contribute to the understanding of digital art as a credible informational resource in the study of the Arts and Humanities disciplines, by reviewing how this material is perceived and supported in the Digital Library environment. A general conclusion is that existing inclusion of digital art in online repositories is predominantly ad-hoc, rather than being the product of an organised effort.
• The project will highlight the particular characteristics of digital art as a representation of digital art objects and display the requirements for description, cataloguing and retrieval that have been previously overlooked. Based on the review of the relevant literature and the findings of the user needs assessment, this work aims to provide a set of concrete recommendations on how to approach this new medium and consequently cater for its long-term preservation.
• The clear focus on assessing and analysing the needs of a specified user community emphasises the imperative role of user involvement in the design of services that encapsulate the expected functionality and help constitute Digital Libraries as creditable resources for learning and research, especially when compared to traditional libraries where a long tradition and experience in accommodating such needs has resulted in a highly-esteemed level of services.
• The proposed solution shows that digital art repositories can benefit from a distributed architecture, in which the material featured in their collections can be conceptualised as a valuable dataset rather than an abstract and arbitrary aggregation of random digital objects.

The methodology presented in this paper is a promising first step in incorporating digital art artefacts in the collections of organised repositories. Although the user needs assessment as designed for the purposes of the project cannot be applied in the context of other endeavours, the patterns that will be highlighted after the analysis of the findings can be used as a starting point in similar projects. A major contribution of the presented approach is that it raises an awareness towards the rationale that DL research should focus on delivering results in narrow domains [22], taking into account the particular requirements of new and emergent digital objects that cannot always be represented or distributed in printed formats [6]. The service-oriented solution proposed in this paper enables a flexible management of information dissemination that is based on reusable, interoperable and dynamically generated applications that can arguably provide cost-effective alternatives to the currently employed three-tier architecture. Nevertheless, this suggestion assumes the cooperation of on-line repositories. In the case of digital art repositories, the collections of primary material are commonly assembled by users and creators with little knowledge of data management, and will most probably remain known and therefore accessible to the microcultures from which they originate. Reaching these repositories can well be a matter of chance or strenuous searching. It is exactly these cases that should be inculcated with the notion of the Digital Library as organised and managed information resources.

Notes

1	The calculations were based on figures provided by the IIESA 2004-2005 statistics (http://www.hesa.ac.uk/holisdocs/pubinfo/student/disab0405.htm) and the Scottish Executive, Students in Higher Education at Scottish Institutions 2004-2005 (http://www.scotland.gov.uk/Publications/2006/04/28100117/0).
2	The case study examines the following cases: The Perseus Digital Library The Library of Congress National Digital Library Program The Alexandria Digital Library (ADL) The Alexandria Digital Earth Prototype (ADEPT) The Desktop Link to Virtual Engineering Resources (DELIver) The Library and Information Service on-line resource of the University of Patras TheGlasgowStory Project The National Digital Library (NSDL)

References

[1]	Aldridge, A. and Levine, K. (2001) Surveying the social world: Principles and practice in survey research. Buckingham: Open University Press.
[2]	Appadurai, A. (2003) Archive and Aspiration. In Aden Mulder et al. (Eds.) Information is Alive - Art and Theory on Archiving and Retrieving Data. Rotterdam: V2_/NAi publishers.
[3]	Bainbridge, D., Nevill-Manning, C.O., Witten, I.H., Smith, L.A., McNab, R.J. (1999) Towards a digital library of popular music. Proceedings of the fourth ACM conference on Digital libraries, p.161-169.
[4]	Bishop A.P., Neumann L.J., Star S.L and Merke C. et al (2000) Digital libraries: situating use in changing information infrastructure. Journal of American Society for Information Science, 51(4).
[5]	Candela, L. et al. (Eds.) (2006) The Digital Library Manifesto, Delos Network of Excellence on Digital Libraries.
[6]	Drabcnstott, K. M. (1994) Analytical Review of the Library of the Future. Washington DC: Council on Library Resources.
[7]	ERCIM (2002) Digital Libraries: Future Research Directions for a European Research Programme. Workshop report 02/W02.
[8]	Hanlon, A. and Coperland, A. (2001) Using the Dublin Core to document digital art : A case study. Journal of Internet Cataloguing, 4(1).
[9]	Jarvis, JD. (2004) Toward a Digital Aesthetic. The Museum of Computer Art.
[10]	Khosrow-Pour, M. (2007) Dictionary of Information Science and Technology. Hershey, PA: Idea Group Reference.
[11]	Kling, R. (2000) Learning about information technologies and social change: The contribution of social Informatics, Information Society, 16(3).
[12]	Kyrillidou, M. (2002) From input and output measures to quality and outcome measures, or, from the user in the life of the library to the library in the life of the user. Journal of Academic Librarianship, 28(1).
[13]	Manovich, L. (2002) Generation Flash. On-line: www.manovich.net/DOCS/generation_flash.doc
[14]	Marchionini G., Plaisant C., and Komlodi A. (2003) The people in digital libraries: Multifaceted approaches to assessing needs and impact. In A. Bishop, B. Buttenfield and N. Van House (Eds.) Digital Library Use: Social practice in design and evaluation. Cambridge: MIT Press.
[15]	Marchionini G., Plaisant C., Komlodi A. (1998) Interfaces and tools for the Library of Congress National Digital Library Program. Information Processing & Management, 34(5).
[16]	McCullough D. (1998) Web-based market research ushers in a new age. Marketing News, 32.
[17]	Minibayeva, N. and Dunn, J.W. (2002) A digital library model for music. Proceedings of the Second ACM/IEEE-CS Joint Conference on Digital Libraries, 154-155.
[18]	Monopoli M., Nicholas D. Gcorgiou P. and Korfiati M. (2002) A user-oriented evaluation of digital libraries: case study the "electronic journals" service of the library and information service of the University of Patras, Greece. Aslib Proceedings, 54(2).
[19]	Plaisant, C., Marchionini, G., Bruns, T.. Komlodi, A.,and Campbell, L. (1997) Bringing treasures to the surface. Iterative designs for the Library of Congress National Digital Library Program. Proceedings of ACM CH197, ACM Press.
[20]	Premkumar, G.P. (2003) Alternate distribution strategies for digital music. Communications of the ACM, 46(9).
[21]	Reviere, R. (1996) Needs Assessment: A Creative And Practical Guide For Social Scientists. Bristol: Taylor & Francis.
[22]	Ross, S. et al. (2003) Invest to Save. Report and Recommendations of the NSF-DELOS Working Group on Digital Archiving and Preservation. On-line: http://delos-noe. iei.pi.cnr.it/activities/internationalforum/JointWGs/digitalarchiving/Digitalarchiving.pdf
[23]	Van House N.A., Butler MIL, Ogle V., Schiff L. (1996) User-Centered Iterative Design for Digital Libraries: The Cypress Experience. D-Lib Magazine, February 1996.
[24]	Waters, D. (2006) Middle East Digital Libraries: Some Principles for Development. Middle East Digital Library Workshop.
[25]	Watson-Verran, H. & Turnbull, D. (1995) Science and other indigenous knowledge systems. In S. Jasanoff, et al. (Eds.) Handbook of science and technology studies. Thousand Oaks CA: Sage.
[26]	Zhang Y. (1999) Using the Internet for survey research: A case study. Journal of the American Society of Information Science, 51(1).