Resumen
Increasing sources and amounts of information challenge users around the globe. Newspapers, television news broadcasts, Web sites and other forms of communication provide society with the vast majority of real-time information. Unfortunately, cost and time pressures demand that producers, editors and writers select and organize content for stereotypical audiences. The article illustrates how content understanding, user modeling and tailored presentation generation offer the potential of personalized interaction on demand. It describes two systems that represent components necessary for an adaptive question answering system, a nonadaptive question answering system and another for tailored news video retrieval. They are investigating systems that will allow users to perform searches by asking questions and getting personalized answers rather than by typing in keywords and typically getting an overwhelming number of often irrelevant Web pages. This requires advancing beyond the conventional information retrieval strategy of document/Web page access toward the automated extraction and summarization of possibly multilingual and multimedia information from structured, unstructured and semistructured sources followed by the generation of a personalized presentation.