Information Seeking Behaviour in Image Retrieval: VISOR I Final Report

The Visor I Final Report is an investigation of image‐seeking behaviour in a variety of workplace domains undertaken in order to increase the understanding of how and why people search for and use visual information. The purposive sampling of real image seekers in museums, libraries, architecture, broadcasting, newspaper publishing, police, medical and art history environments provided both depth and breadth to the issues under investigation. The authors should be commended on their ability to take a panoply of subject matter details and provide concise, yet revealing, insights into the realm of image retrieval from a user‐centred perspective. In addition, the interdisciplinary facets of this study provide a rich resource for further research, not only in the area of human behaviour during image seeking and retrieval, but in adjunct frameworks grounded in communication theory, cognitive science, organisational development and knowledge management.

The data were collected from 45 users of both manual and digital image retrieval systems in ten different organisations. With no coherent body of knowledge in existence, data collection was triangulated through the use of interviews, verbal protocols, questionnaires, informal observation and organisational documentation. The approach to analysis was grounded strictly in an a posteriori examination of the data with no preconceived notions or theoretical perspectives. If there is any caveat surrounding the results and modelling of the human interactions described in this report, it could stem from the construct validity associated with the algorithmic authority imposed by the software package (Atlas/ti) used to analyse the interview and verbal protocol transcriptions.

The reported findings are in a format derived from three approaches designed to:

1. (1)

   encapsulate the level and depth of the data;

2. (2)

   define users’ behaviours in context; and

3. (3)

   depict interactional relationships between factors.

Accuracy was mediated through user self‐reporting, researcher observations and user feedback/verification of an initial draft report. The value of the reported results is engendered by the form of presentation, which included an overview of each organisation involved in the study, the different kinds of image use investigated, the types of image search conducted and the mapped relationships of users to their various image‐seeking environments.

Although the brevity of this review prohibits a lengthy discussion of the findings, it is important to note the following as encouragement to future readers:

• •

  Users appear to invoke a behavioural process for seeking images that embrace: starting, scoping, applying, selecting, iterating and ending.

• •

  Image‐seeking behaviour normally occurs in a contextual framework that includes five key categories: workplace, resources, personal influences, communications/interactions, and access mechanisms. Each of these categories can have a mediating effect on understanding how and why people seek, retrieve and use visual information.

• •

  The inductively inferred was deductively confirmed that time and money are major factors that influence the success of human behaviour in image‐seeking and retrieval environments.

While some might criticise the reported results as mostly statements of the intuitively obvious, the qualitative rigour employed to compile, analyse and model human interactions in a variety of organisational settings makes this report a must read for anyone involved in image seeking, retrieval and use from a user‐centred viewpoint.