This is an archived page. Visit the current R&D website at

Kepler (2013)

Since 1913, The New York Times has been publishing an index of its articles, organized by the places these articles took place, the people these articles covered, the concepts and issues these articles addressed, and other facets. In effect, The Times was publishing metadata before the term came to its modern use.

As we investigate new ways to understand the volume of information The New York Times produces every day, we often use these tags as a tool to see how articles relate to each other, to see what and who is being covered, and to find unique and interesting combinations of topics. By combining the tags on each article with the viewership the article receives, we can interpret interest among our readers at a subject level that surpasses a single article or column. And by looking at the topics for those articles that are shared on Twitter, we can get a sense for those subjects that generate the most conversation and distribution.

Kepler is a visualization of these three data sources as they relate to each other. Connections drawn between words indicates an article that includes both connected concepts. Stories are shown as stars in these constellations, and both words and stories are sized according to one of three metrics: words written, page views recorded, or Tweets observed. This particular application is designed for a more ambient display, but has inspired us to look at ways that better use of topic can lead to a more exploratory navigational experience through the day's news.