The topic normalized impact factor

Introduction Traditionally, normalization for field differences has usually been done based on a field classification system. In said approach, each publication belongs to one or more fields and the citation impact of a publication is calculated relative to the other publications in the same field. An example of a field classification system is the JCR subject category list. For these subject categories, Egghe & Rousseau (2002) propose the aggregate impact factor in a similar way as the JIF, taking all journals in a category as one meta-journal. However, the position of individual journals of merging specialties remains difficult to determine with precision and some journals are assigned to more than one category. In this sense, Dorta-González & Dorta-González (2013a) propose the categories normalized impact factor considering all the indexing categories of each journal. Recently, the idea of source normalization was introduced, which offers an alternative approach to normalizing field differences. In this approach, normalization is achieved by looking at the referencing behaviour of citing journals. Many indices, such as the fractionally counted impact factor (Leydesdorff & Bornmann, 2011), dividing each citation by the number of references, and the 2-year maximum journal impact factor (DortaGonzález & Dorta-González, 2013b), considering the 2-year citation time window of maximum impact instead of the previous 2-year time window, have been proposed. However, all these metrics do not include any great degree of normalization in relation to the specific topic of each journal. In this sense, we use the aggregate impact factor of the citing journals as a measure of the citation potential in the journal topic, and we employ this citation potential in the normalization of the journal impact factor to make it comparable between scientific fields.