Improving Information Retieval by Modelling Business Context

Retrieving information is a central operation when accomplishing most of today business tasks. Unfortunately, a person has hard time finding relevant information necessary to accomplish its business task due to the complexity of these information processes. In certain corporate bodies (aeronautical, automotive, etc.) these tasks are critical and should be run from specific information. As a solution, information retrieval systems must take into account the user context which integrates the user business task to improve its overall accuracy and facilitate the achievement of the task. To solve this problem we present in this paper the architecture of an information retrieval system applicable to a business situation. It is based on a triptych context model which integrates three dimensions: user, task and environment. This model is associated with a learning process and to a process that creates according to the conjunction of all the contextual factors, a particular situation. This situation will be exploited by the information search system to improve its accuracy in being the entry point of the information retrieval system. We explain how this system can improve the information search in a business context by taking into account cognitive, technical and social aspects.