In the original article, Equation (1) in Appendix B in Data Sheet 1 contains an error. The correct equation is:
(1) mean[GNsim(word, label_1pos) + … + GNsim(word, label_Npos)] − mean[GNsim(word, label_1neg) + … + GNsim(word, label_Nneg)]
where GNsim is the so-called Lin similarity (Lin, 1998) defining semantic relatedness via a formula derived from information theory. This measure is sometimes called a universal semantic similarity measure as it is supposed to be application-, domain-, and resource independent (cf. Budanitsky and Hirst, 2006).
label_1pos and label_1neg/label_Npos and label_Nneg are the first and last terms, respectively, in either the valence or AP lists given in S2 and S3 of the supplementary materials, i.e., BEFRIEDIGUNG (satisfaction), ANGST (fear), or VERGNÜGEN (have fun), TRAUERN (mourn), and ANMUT (grace), WONNE (delight), or ABSCHEU (abomination), ZUMUTUNG (impertinence).
The original file Data Sheet 1 in the Supplementary Material has been updated.
Conflict of interest statement
The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
References
- Budanitsky A., Hirst G. (2006). Evaluating wordnet-based measures of lexical semantic relatedness. Comput. Linguist. 32, 13–47. 10.1162/coli.2006.32.1.13 [DOI] [Google Scholar]
- Lin D. (1998). An information-theoretic definition of similarity, in Proceedings of the Fifteenth International Conference on Machine Learning (ICML'98) (Madison, WI: ), 296–304. [Google Scholar]