Taxonomy of personalisation for Generating personalised content in Technical Support Forums

There are two traditional approaches to meeting international users' requirements during content generation. The first is localisation which adapts a content to meet the language, cultural and other requirements of a specific locale. The second is personalisation which involves delivering relevant content and presenting information according to individual users’ preferences. The fundamental question that must be answered in generating personalised content is: what are the relevant attributes for personalising content? Work to date in personalisation has been based on several logicbased standards and frameworks that have been proposed. However these different standards have led to a degree of inconsistency in the field and are open to accusations of validity. This research aims to empirically identify the relevant attributes for personalising content. It focuses on data obtained from technical support forums on the web, a growth area in terms of customer support. It uses a grounded-theory based approach to analyse the conversations on these forums in order to identify the personalisation attributes of the support provided. This paper reports on a preliminary study in this work, which analyses data from a number of technical support forums, and presents an initial taxonomy of empirically derived personalisation issues for this domain.