Who gets acknowledged: measuring scientific contributions through automatic acknowledgment indexing

Abstract

Acknowledgements in research publications, like citations, indicate influential contributions to scientific work. However, acknowledgements are different from citations; whereas citations are formal expressions of debt, acknowledgements are arguably more personal, singular, or private expressions of appreciation and contribution. Furthermore, many sources of research funding expect researchers to acknowledge any support that contributed to the published work. Just as citation indexing proved to be an important tool for evaluating research contributions, we argue that acknowledgements can be considered as a metric parallel to citations in the academic audit process. We have developed automated methods for acknowledgment extraction and analysis and show that combining acknowledgment analysis with citation indexing yields a measurable impact of the efficacy of various individuals as well as government, corporate, and university sponsors of scientific work.

Fig. 1.

The proportion of all documents indexed by DBLP that are contained in CiteSeer by year.

Fig. 2.

The proportion of all publication venues in DBLP contained by CiteSeer, where the venues are ordered by the amount of coverage received in CiteSeer.

Fig. 3.

The distribution of acknowledgments in the CiteSeer document collection follows a power law with the exponent –0.65. A line with –0.65 slope is drawn for reference.

Fig. 4.

The proportion of documents per year acknowledging the 10 most acknowledged entities in the CiteSeer document collection.