How many keywords do authors assign to research articles – a multi-disciplinary analysis?

Author keywords are one important data source for co-word analysis. The distri-bution of author keywords in papers has not been investigated at the discipline level. We analyzed six research fields from soft science to hard science to reveal the underlying quantitative patterns of author keywords. Normal distribution, Poisson distribution, and Weibull distribution were fitted by applying Maximum Likelihood Estimation. Chi-Square tests and Kolmogorov-Smirnov tests were used to evaluate the goodness of fit. The results show that a large portion of pa-pers have no keyword or only one keyword in all these fields. The author key-word distributions of the six fields are represented. It’s shown that Weibull dis-tribution is the best fitted. This study provides practical implications for keyword selection in co-word analysis