Odysseus

University College Dublin holds a variety of primary source repositories within its confines. Of major importance are the holdings of the School of History and Archives, the School of Irish, Celtic Studies, Irish Folklore and Linguistics, and the Library's Special collections. The repositories contain physical materials in documentary, audio, video and graphic formats. Their uniqueness or rarity, and in many cases their fragility, is a major obstacle to their availability and use by the scholarly community outside of UCD, while within the university, their physical dispersal is counter-productive to their effective use. The Irish Virtual Research Library and Archive (IVRLA) project aims to realise the latent potential of these collections by: (1) creating critical mass of digital objects representing all the main media types held in UCD, and (2) exploring innovative ways of accessing and adding value to these digital objects using a variety of technical and pedagogical means. As part of their agenda The Humanities Institute of Ireland (based in UCD) have sought the collaboration of the School of Computer Science and Informatics at UCD in relation to the latter of these goals.

This project is part of a broader collaboration with the IVRLA project, and will investigate ways to help researchers to identify, locate and retrieve digital objects of relevance to their areas of interest or study regardless of media type or relevance. In particular, for years 1 & 2 of this project the intention is to demonstrate how the reachability of this archive can be extended, and to investigate how best to recognize relationships between and amongst currently disparate archival collections, through the integration of collaborative search techniques and sophisticated recommendation technologies. Currently, a student of Irish folklore may consult very different references within the "Irish Folklore Collection", to other students that study in this same area, and are often ignorant to (and geographically distant from) a wealth of relevant resources that are contained within "The Irish Folklore" archive, for instance.

The proposal is to recruit 1 student to work on this project. In this first Summer they will work (in collaboration with the Humanities School) on the Data Collection and Representation task, and implementation of the core Recommendation and Search Engine component. The idea is to gather user search and browsing information over a nine-month period using a portion of the digitised content available at that time. Next year the intention would be to use the architecture and user logs from year one of the project to investigate more sophisticated recommendation and search approaches, and evaluate the performance of these, once again, in the context of real-user studies. Another key research question to be addressed by the project is the ability of digital archives to be used collaboratively and in a teaching environment with such concepts as electronic acetate allowing for a researcher or lecturer to annotate materials for interested parties.

Relevance of Project to the Host Laboratories:

IVRLA is an exciting project that will develop and mature into a valuable service for UCD students and academics, in addition to enabling nationwide and international access. The project will result in vulnerable unique primary source material being available to browse and search in an on-line environment. Also, the opportunity to collect and analyse real-user data of this nature is rare, and having access to this resource will greatly facilitate the development of more sophisticated search and recommendation strategies that are inspired by the actual behaviours of those users.

Supervisors:

Dr. Jill Freyne and Dr. Lorraine McGinty (Computer Science and Informatics, UCD)

Keywords:

Search and Recommendation, Digital Achieving, Analysis of User Behaviour.