ePaper

Personalized Newspaper

Competence Center:  Information Retrieval and Machine Learning
ContactProf. Dr. Sahin Albayrak

Partners: Deutsche TelekomdpaFraunhofer FIT

The goal of the ePaper project is the development of a personalized newspaper application for mobile devices that provides the same intuitive handling as printed media with enhancements derived from the capabilities of the devices used. Benefits of ePaper include being handier than carrying a number of magazines and newspapers, offering personalized content compilation and interactivity. ePapers integrate multimedia enabled content which enables new user experiences and allow the access to indexed local and remote archive of issues and articles. Furthermore, the consolidation of information from different sources enables comprehensive overview across content providers. 

Overview

The ePaper application provides a fresh, personalized newspaper every day by collecting and displaying news articles from different sources. The main research focus is on the automatic production of page layouts that resemble the look&feel of traditional printed newspapers.

The prototype additionally contains a user modeling component that allows for the long-term tracking of user interactions and interest. This implicit user feedback is used to personalize news article selection and layout. For example, news articles deemed very interesting for the user can be placed early in the newspaper and in prominent positions, whereas less important articles are put on less space. This is partly achieved through an automatic condensation of news articles using text summarization technologies.