conference2009@iskouk.org

Poster Abstracts

This presentation illustrates a framework for analysing the objectivity of taxonomy projects and shows how it can illuminate the political nature of practical taxonomy work as mediation of differing viewpoints, or as a balancing of subjectivity and objectivity.

Theoretical background. The US philosopher Helen Longino proposes that the subjective/objective distinction is a false dichotomy in a scientific inquiry, arguing instead that objectivity depends on a process of intersubjective creation of meaning (Longino, H. (1990) Science as Social Knowledge). Similarly the creation of a taxonomy depends on negotiating an agreement on terminology choices and categorisations within a particular socio-cultural context. Longino asserts that for scientific inquiry to be objective, it must satisfy four criteria:

• Openness to criticism
• Responsiveness to criticism
• Public accessibility of standards
• Equality of intellectual authority of contributors

These criteria can be used as a framework for assessing the essentially “political” process of mediation that a taxonomist must undertake, illuminating the degree of “objectivity” of a taxonomy project.
Methodology and results. Fourteen taxonomy professionals were interviewed about 15 diverse taxonomy projects and their responses scored against a set of five questions exploring each of Longino’s four criteria. The projects were ranked. Large-scale public projects scored higher than projects in the commercial sector and small-scale or specialist projects scored lowest.

Conclusions. This research intends to show that Longino’s framework is a useful way of highlighting the degree of “objectivity” of a taxonomy project, and that the level of “objectivity” is related to the political nature of the project and how the role of “taxonomist as politician” is formally supported. There would also seem to be a similarity between established industry best practice (especially for well-resourced projects), and satisfaction of Longino’s criteria for objectivity.
[Short paper]

The central aim of this poster is to describe the development of a virtual resources library with specialized contents on the field of cardiovascular diseases, organized according to a faceted organisation of categories. We understand “facets” as the classes of the different categories of one specific subject field. The information of our e-library is classified following two main different criteria: On one hand depending on the subject(s) of the documents, ordered by an onomasiological structure. For instance, we sorted the cardiovascular diseases into more specific subfields like cardiovascular abnormalities (vascular malformations, heart defects, etc.); heart diseases (arhythmias, heart failure, heart neoplasms, etc.); vascular diseases (hypertension, stroke, etc.).

On the other hand we take into consideration the characteristics and attributes of the more usual information in the biomedical areas. The classification was made depending on the form, the structure or the type of content of the document and we had classes like reference documents (atlas, books, clinical practice guidelines, databases, dictionaries and glossaries, etc.); information about congresses and other meetings; health and medical portals or associations. In addition, all the different categories and classes (subjects as well as document types) have been labelled in two languages, Spanish and English, to allow the recovery of the information no matter which language is used.

Without a doubt one of the most important aspects for the quality of information resources is the accuracy of the retrieval of the contents, avoiding the silence and the documental noise and more precisely the faceted classification methodology gives us the possibility to deal with more than one category and so increasing the relevancy of the search results.

The EASAIER system has been developed by a pan-European consortium to archive audio and related materials, and allow users to interact with the materials both on the web and via client software. A range of existing archives have been involved in the development as an ‘Expert User Advisory Board’ to inform the developers of their various technical and information needs so these may be accommodated in the software design, using an iterative approach. Some also installed elements of the system within their archives as part of the deployment testing and evaluation. The range of types of formats, size of collections, metadata standards, type of technological infrastructure and end user and expert user needs and requirements are all important factors in successful deployment of novel systems. These are compounded by the special nature of multimedia. This poster discusses the archives’ needs in relation to the client, archiver and server’s performance, usefulness and usability, and presents an integrated approach to evaluation and deployment which may be used to inform development of similar systems. Specific problems that were found during test deployment included inaccurate documentation, coding errors, bugs, firewalls, a wide range in technical abilities of archive staff, and lack of internal IT support. The strength of the deployment team is shown to be in its response to these problems, in an attempt to successfully deploy the system.

This poster shows how a metadata model can be the backbone for a large information and content-rich government site, known as Directgov. The aim of the project was to demonstrate how a metadata-driven paradigm could improve the findability and usability of content, and provide new ways for content to be aggregated in useful and intuitive ways. The model was designed as part of a user-centred design approach, and we worked alongside user experience designers to create a model that would help support rich user journeys, as well as improve the efficiency of editorial processes that are currently manual and time-intensive.

Directgov’s vision is to be the citizen-focussed digital channel for government. By 2011, Directgov will be the principal online destination for citizens to interact with government online. It already receives over 18 million visits a month, and brings together information from all over government. The convergence programme is reducing the number of government websites, and bringing their content onto Directgov. This means the site will grow in size and complexity, hence the need for a robust, scaleable information architecture. The metadata model is a core part of this.

The model and its associated taxonomies provide consistent, user-friendly metadata terms to build applications on, for example, powering facets and filters in Search, and supporting a “What’s new?” aggregation of the most recent content to be published.

The model has to meet both end-user and business needs in four key areas:

• Describing content
• Enabling Search enhancement
• Managing content
• Specifying ownership of content

Each of these areas is illustrated, showing some of the applications and the metadata fields that will enable them. Our work shows how taxonomies and structured metadata are part of the cutting-edge thinking powering the joining-up of government information across departments.

The poster explores the University of Melbourne Archive's management and facilitation of access to photographic, objects, and paper based collections; addressing issues of hosting and metadata mapping; and including digitization, migration, access, facilitation of external harvesting, and metadata development and maintenance within a complex architectural environment involving a number of platforms, databases, software and systems. The poster includes:

• The origins and design of several collection databases (essentially metadata repositories)
• Australian and international initiatives and progress towards management of Digital Collections
• Data Migration initiatives – progress and pitfalls - where to from here? (trying to fit a square peg in a round hole)

The final goal of Information Retrieval (IR) is knowledge production. However, it has been argued that knowledge production is not an individual effort but a collaborative effort. Collaboration in information retrieval is geared towards knowledge sharing and creation of new knowledge by users. While many attempts have been made in expert systems for capturing knowledge and reusing these knowledge, we discovered that their response is based on the pattern coded in the system whereas human beings, influenced by so many factors, possess some tacit knowledge which they themselves may not be aware of until they are faced with problems that will steer up that knowledge in them. This paper aims at explaining how users’ tacit knowledge can be made explicit through collaboration during IR by allowing synchronous human to human interaction in information retrieval through the mediation of a Collaborative Information Retrieval System (CIRS) and through annotation of the objects in the CIRS thus enhancing knowledge creation and structuring.

Many repositories aggregate resources from metadata and are applied in many industries. In this paper, I try to build a folksonomies evolution architecture from those metadata tags. The folksonomies architecture will generalize to many fuzzy ontologies. Based on the fuzzy ontologies, this paper will also show how those metadata transfer to a flexible and automatic semantic retrieval service.
[Poster outline]

For enquiries please contact: conference2009@iskouk.org