← Back to introduction

trend h-index

The trend h-index (Sidiropoulos et al. 2007) is essentially the opposite of the contemporary h-index. It is designed to measure how current an author's impact is by how recently they are being cited. The trend score for a publication is measured as

$$S^t_i = \gamma \sum\limits_{j=1}^{C_i}{\left(Y-Y_{j.i}+1\right)^{-\delta}}$$

where γ and δ are parameters (often set to 4 and 1, respectively, just as with the contemporary h-index) and Yj.i is the year of the jth citation for publication i. If the number of citations for publication i in year Yk is Ci.k, this can also be written as $$S^t_i = \gamma \sum\limits_{k=1}^{Y}{C_{i.k}\left(Y-Y_k+1\right)^{-\delta}}.$$The trend h-index is the largest value for which an author has ht publications with at least Stht.

$$h^t = \underset{i}{\max}\left(i \leq S^t_i\right)$$

Example

Publications are ordered by adjusted number of citations, from highest to lowest.

Citations (Ci)592616111110342111300000
—1997—100010
—1998—63310100
—1999—12102121000
—2000—134645602001000
—2001—2795742312111000000
Adjusted Citations (\(S^{t}_i\))156.8060.3337.6736.0028.3322.6712.009.338.004.004.004.003.800.000.000.000.000.00
Rank (i)123456789101112131415161718
hC = 8

The largest rank where \(i \leq S^{t}_i\) is 8.

History

Yearht
19972
19984
19995
20007
20018
200211
200314
200418
200521
200621
200723
200826
200927
201031
201133
201233
201334
201434
201535
201636
201736
201836
201938
202038
202139
202237
202336
202438
202537

References