Pudovkin, A.I.; Garfield, E.: Algorithmic procedure for finding semantically related journals (2002)
0.01
0.01102142 = product of:
0.02204284 = sum of:
0.02204284 = product of:
0.04408568 = sum of:
0.04408568 = weight(_text_:j in 5220) [ClassicSimilarity], result of:
0.04408568 = score(doc=5220,freq=6.0), product of:
0.14500295 = queryWeight, product of:
3.1774964 = idf(docFreq=5010, maxDocs=44218)
0.045634337 = queryNorm
0.304033 = fieldWeight in 5220, product of:
2.4494898 = tf(freq=6.0), with freq of:
6.0 = termFreq=6.0
3.1774964 = idf(docFreq=5010, maxDocs=44218)
0.0390625 = fieldNorm(doc=5220)
0.5 = coord(1/2)
0.5 = coord(1/2)
- Abstract
- Journal Citation Reports provides a classification of journals most heavily cited by a given journal and which most heavily cite that journal, but size variation is not taken into account. Pudovkin and Garfield suggest a procedure for meeting this difficulty. The relatedness of journal i to journal j is determined by the number of citations from journal i to journal j in a given year normalized by the product of the papers published in the j journal in that year times the number of references cited in the i journal in that year. A multiplier of ten to the sixth is suggested to bring the values into an easily perceptible range. While citations received depend upon the overall cumulative number of papers published by a journal, the current year is utilized since that data is available in JCR. Citations to current year papers would be quite low in most fields and thus not included. To produce the final index, the maximum of the A citing B value, and the B citing A value is chosen and used to indicate the closeness of the journals. The procedure is illustrated for the journal Genetics.
