User perceptions of relevance and its effect on retrieval in a smart textile archive.

Horsburgh, Ben; Craw, Susan; Williams, Dorothy; Burnett, Simon; Morrison, Katie; Martin, Suzanne

Abstract

The digitisation of physical textiles archives is an important process for the Scottish textiles industry. This transformation creates an easy access point to a wide breadth of knowledge, which can be used to understand historical context and inspire future creativity. The creation of such archives however presents interesting new challenges, such as how to organise this wealth of information, and make it accessible in meaningful ways. We present a Case Based Reasoning approach to creating a digital archive and adapting the representation of items in this archive. In doing so we are able to learn the important facets describing an item, and therefore improve the quality of recommendations made to users of the system. We evaluate this approach by constructing a user study, which was completed by industry experts and students. We also compare how users interact with both an offline physical case base, and the online digital case base. Evaluation of our representation adaptation, and our comparison of physical and digital archives, highlights key findings that can inform and strengthen the process for creating new case bases.

Citation

HORSBURGH, B., CRAW, S., WILLIAMS, D., BURNETT, S., MORRISON, K. and MARTIN, S. 2013. User perceptions of relevance and its effect on retrieval in a smart textile archive. In Delany, S.J. and Ontañón, S. (eds.) Case-based reasoning research and development: proceedings of the 21st International conference on case-based reasoning (ICCBR 2013), 8-11 July 2013, Saratoga Springs, USA. Lecture notes in computer science, 7969. Berlin: Springer [online], pages 149-163. Available from: https://doi.org/10.1007/978-3-642-39056-2_11

• Conference Name: 21st International conference on case-based reasoning (ICCBR 2013)
• Conference Location: Saratoga Springs, USA
• Start Date: Jul 8, 2013
• End Date: Jul 11, 2013
• Acceptance Date: Jul 11, 2013
• Online Publication Date: Jul 11, 2013
• Publication Date: Dec 31, 2013
• Deposit Date: Dec 10, 2014
• Publicly Available Date: Dec 10, 2014
• Print ISSN: 0302-9743
• Publisher: Springer
• Pages: 149-163
• Series Title: Lecture notes in computer science
• Series Number: 7969
• Series ISSN: 0302-9743
• ISBN: 9783642390555
• DOI: https://doi.org/10.1007/978-3-642-39056-2_11
• Keywords: Learning refined representation; Digitisation of physical archives; User evaluation
• Public URL: http://hdl.handle.net/10059/1094