Abstract
The Chandra Data Archive (CDA) maintains an extensive observatory bibliography. By linking the published articles with the individual datasets analyzed in the paper, we have the opportunity to join the bibliographic metadata (including keywords, subjects, objects, data references from other observatories, etc.) with the meta- data associated with the observational datasets. This rich body of information is ripe for far more sophisticated data mining than the two repositories (publications and data) would afford individually. Throughout the course of the mission the CDA has investigated numerous questions regarding the impact of specific types of Chandra programs such as the relative science impact of GTO, GO, and DDT programs or observing, archive, and theory programs. Most recently the Chandra bibliography was used to assess the impact of programs based on the size of the program to examine whether the dividing line between standard and large projects should be changed and whether another round of X-ray Visionary Programs should be offered. Traditionally we have grouped observations by proposal when assessing the impact of programs. For this investigation we aggregated observations by pointing and instrument configuration such that objects observed multiple times in the mission were considered single observing programs. This change in perspective has given us new ideas for assessing the science impact of Chandra and for presenting data to our users. In this paper we present the methodologies used in the recent study, some of its results, and most importantly some unexpected insights into assessing the science impact of an observatory.
