A Mid-term Review Paper Prepared

Introduction Categorisation is something that we do naturally and unconsciously every day. We recognise one animal as a cat and another as a dog. We organise objects in the world around us in ways that reflect these categories. In our kitchens, we keep baking trays with other baking trays, saucepans with other saucepans and keep food separate from cleaning products. We categorise ideas, people, tasks and objects. Categorisation is fundamental to the way we think. Yet when we categorise things in information systems, problems frequently occur. The folders in our personal computer systems are often disorganised and messy. Databases seem to be full of entries labelled 'other'. The entry we are trying to file just does not seem to fit nicely into any of the available categories. If we categorise so naturally in our brains, why is it that our information systems are so disorganised? In industries where knowledge is essential to maintaining competitive advantage, a disorganised information system can severely hamper workers. What good is a knowledge base if I cannot find the information I need within it? 2 The research described in this paper explores why this disorganisation is so prevalent and what we can do to minimise it. Beginning with a motivating study, we examine some of the common categorisation problems that arise in information systems supporting knowledge management. We then move on to explore some of the cognitive reasons why these problems occur. Following that we explore problems arising from the context in which categorisation occurs, such as political and social causes. Based on this understanding of the different causes of categorisation problems, we then define our research problem. Finally, we propose avenues of research to address this.