Empowering web portal users with personalized text mining services

Nowadays, many organizations use portals extensively as a single-point access to information, applications, and people. However, dealing with the constantly growing amounts of information available through web portals is difficult and time-consuming for users. Most of the current portal systems enable users to retrieve content statically defined as relevant - but reading and interpreting it remain a serious bottleneck. We propose to break this bottleneck with a personalized information system that integrates Natural Language Processing (NLP) to support users in analysing, transforming, and creating knowledge from large amounts of textual content. Our approach is a novel combination of web portal technology with the Semantic Assistants (SA) framework (Witte and Gitzinger, 2008), an extensible software architecture that allows invoking literally any NLP or text mining tool using either Web Services or Application Programming Interfaces (API). The whole system is designed to give users full control over personalization, and leverage visualizations to adjust the adaptive behaviour to the users’ preferences in an easy-to-use way.