Web analytics as tool for improvement of website taxonomies

The poster examines how web analytics can be used to provide information about users and inform design and redesign of taxonomies. It uses a case study of the website Cancer.dk by the Danish Cancer Society. The society is a private organization with an overall goal to prevent the development of cancer, improve patients’ chances of recovery, and limit the physical, psychological and social side-effects of cancer. The website is the main channel for communication and knowledge sharing with patients, their relatives and professionals. The present study consists of two independent analyses, one using Google analytics focusing on searching and browsing activities, another using a home-grown transaction log developed to collect data about tagging, searching and browsing by tags. The log is set up to distinguish between tags added by editors and end-users respectively. Altogether, the study provides information about e.g. subjects of interest, searching behaviour, browsing patterns in website structure as well as tag clouds, page views. The poster discusses benefits and challenges of the two web metrics, with a model of how to use search and tag data for the design of taxonomies, e.g. choice of categories and vocabulary, hierarchical structure, granularity, associative navigation, and filtering.