# Evaluative Bibliometrics is on-line

Morris, Steven (BA) Steven.Morris at BAKERHUGHES.COM
Sun Mar 9 11:14:12 EDT 2008

```
Ting,

You can count the entities and links in the database or in DIVA.

In the database:

1) the number of papers is the number of lines in 'working table'
2) the number of references is the number of lines in 'cite_keys'
3) the number of links is the number of lines in 'citations'

For any other paper to entity matrix you'll find the number of entities is the number of keys in the keys table, and the number of links is the number of lines in the paper to entity table.

In DIVA, you have to know the name of the matrix corresponding to what you're interested in.

For example the paper to reference matrix is project.raw.ref_mat
In the command window execute the line:

>> size(project.raw.ref_mat)
ans =
1659       28720

The first number is  the number of papers, the second number is the number of references.

To get the number of links execute the command:
>> nnz(project.raw.ref_mat)
ans =
84419

This number counts the sum of the number of secondary entities linked by the primary entity. For most of the matrices, which are binary matrices, this is the same as the sum of the number of links from each primary entity to each secondary entity, including multiple links.

To get the total number of links you have sum all elements in the matrix using the command:

>> sum(sum(project.raw.ref_mat))
ans =
(1,1)          84419

The paper reference matrix is binary (only zero or 1 link per entity pair), so the both methods give the same answer.

However for the paper journal to reference journal matrix, the numbers are totally different:

>> nnz(project.raw.jp_jr)
ans =
5094

>> sum(sum(project.raw.jp_jr))
ans =
(1,1)          84467

That's because each paper journal cites each reference journal many times and the paper journal to reference journal is not binary.

Got it?

Thanks,

Steve

________________________________

From: ASIS&T Special Interest Group on Metrics on behalf of Francis Narin
Sent: Sat 3/8/2008 11:57 PM
To: SIGMETRICS at LISTSERV.UTK.EDU
Subject: [SIGMETRICS] Evaluative Bibliometrics is on-line

I have just learned that Steve Morris scanned a copy of my 1976 Monograph " Evaluative Bibliometrics : The Use of Publication and Citation Analysis in the Evaluation of Scientific Activity" . It is available at his web site

http://www.conceptsymbols.com/narin/narin_1975_eval-bibliometrics_images.pdf

In looking over the Monograph after 32 years it is surprising how many of the topics we dealt with in 1976 are still being actively discussed today, from adequacy of the SCI, to journal and subfield mapping, to journal influence, scientific productivity, correlation with non-literature methods, university rankings and the like.

Most of you are probably not aware that the Influence Methodology, which is Chapter VII of the monograph, and also found in "Citation Influence for Journal Aggregates of Scientific Publications: Theory, with Application to the Literature of Physics,"Gabriel Pinski and Francis Narin.  Information Process­ing and Management, 12, 5, 297-312, 1976. ,  which Gabe Pinski and I developed to have a better way of ranking  journals than the Impact Factor, was cited and used by  Sergey Brin and Lawrence Page at Stanford in developing the strategy for a new search engine they called  Google. See  http://www.cs.cornell.edu/home/kleinber/sciam99.html

Francis Narin

```