User needs and indexing practices in a digital newspaper photograph archive

Abstract

ABSTRACTThe production and transfer methods of digital images have evolved rapidly and the digital image material in various organizations is growing fast. New information retrieval (IR) solutions are needed to provide access to images. The requirements of the data type, user needs and work tasks and processes need to be studied. Much research focuses on the development of image indexing methods and algorithms. However, the research on user needs seems to come behind the rapid technical development. For images, considerably few user studies exist, on which to found the system design.This work presents a field study on the needs and searching behaviour of journalists and indexing performed by archivists in a digital newspaper photograph archive. Observations and interviews were used to collect data on journalists' illustration tasks, photograph needs and selection criteria applied and on indexing practiced by the archivists including photograph attributes indexed and the specificity of indexing. In addition, a sample of requests given to archivists and a sample of photograph captions produced in photograph production process were collected and analysed.The results obtained on user needs were contrasted with observed indexing practices and problem areas in the end-user access to photographs were identified. The analysis revealed that the users coped reasonably well when searching photographs of named persons or current news events, which needs were also prevailing in the data. Photographs of object types, named places or themes were also needed but considered difficult to find. Browsing was the key searching strategy applied by the users. The access problems faced by the users are discussed and potential solutions by developing the indexing and the user interface are proposed.

Asiasanat:image retrieval, user needs