Tell Me More is a system that scans online news articles and returns paragraphs that contain new information, beyond that reported in a given article. The primary intellectual contribution of this research is the sophisticated multi-criteria measure of new information, used to select supplemental paragraphs to display. In addition to a semantic-based measure of novelty, the system uses references to new actors (people, organizations, and so on), quantitative information not present in the original article, figures, and new quotations concerning the subject matter. Iacobelli, Birnbaum, and Hammond also report on a small study conducted to gauge users’ reactions to their system and its user interface.

This is an example of mature research that incorporates algorithms from a number of sources to synthesize a highly functional system. Systems such as this one will have an important role to play as traditional media moves to produce more online content. This research will be of interest to computational linguists, human factors practitioners, and the developers of online news systems.