The Citation Metrics Report

The Citation Metrics Report is an overview of citations, usage and derived indicators for a set of ADS records. This means that all quantities in this overview are solely based on data from the ADS. For each paper, a “read” is counted if an ADS user runs a search in our system and then requests to either view the paper’s full bibliographic record or download the fulltext. Please note that in computing readership numbers we attempt to remove log entries generated by robots, users coming to an ADS record from an external search engine, and multiple clicks from the same user. The ADS has an automated procedure that attempts to match the references in the bibliography of a record to existing records in the ADS database. If an ADS record has N references associated with it, it means that the corresponding paper has at least N references in its bibliography, but potentially more if there were references that our procedure was unable to match. Keep this in mind when regarding citation and citation-derived information.

How to View Metrics:

• To access the metrics view for a list of results, on the search results page go to the Explore button –> Citation Metrics.
• To access metrics for a single article, go to the article detail view and find the metrics in the left-hand navigation.

Citation Metrics

(Note that we do not remove self-citations based on author name, because of author disambiguation problems. We apply a list-based removal of self-citations.)

Normalized Citations

For a list of N papers (i=1,…N), where Nauthi is the number of authors for publication i and Ci the number of citations that this paper received, the normalized citation count for each article is Ci/Nauthi, and the ‘normalized citations’ for this list of N papers is the sum of these N numbers.

Article Usage Metrics

The ADS collects information about the rate that articles have been accessed in our system. This data includes both short-term access data (recent views) and records of article access rates over the years. Article usage rates include both page views of the article detail page in the ADS, and full-text downloads of the article. The value for just the full-text downloads is presented as a seperate field in the metrics page.

Recent Views

This data encompasses the past 90 days of access data for articles in the ADS, including full-text downloads. This value also includes read and download information from arXiv. You can sort a list of ADS search results by this value, or view it in the results graph visualization.

Metrics for Indices

h-index

Hirsch’s h-index is the largest number H such that H publications have at least H citations. It attempts to measure the productivity and impact of a researcher in a single number. Wikipedia entry

m-index

The m-index is the h-index divided by the time (years) between the first and most recent publication.

iN-index (where N is 10 or 100)

The iN-index is the number of publications with at least N citations.

g-index

Given a set of articles ranked in decreasing order of the number of citations that they received, the g-index is the (unique) largest number such that the top g articles received (together) at least g² citations.

tori-index

The total research impact of a scholar (tori) is calculated using the reference lists of the citing papers, where self-citations are removed. The contribution of each citing paper is then normalized by the number of remaining references in the citing papers and the number of authors in the cited paper. The tori-index is defined as the amount of work that others have devoted to his/her research, measured in research papers (see Pepe & Kurtz 2012).

riq-index

The research impact quotient (riq) equals the square root of the tori-index, divided by the time between the first and last publication, multiplied by 1000 (see Pepe & Kurtz 2012).

read10-index

Read10 is the current readership rate for all an individual’s papers published in the most recent ten years, normalized for number of authors (see Kurtz et al. 2005).

Paper Count Metrics

Normalized paper count

For a list of N papers (i=1,…N), where Nauthi is the number of authors for publication i, the normalized paper count is the sum over 1/Nauthi