The first documents found in the Scopus database was published in 1999. It was written by A. Morrison with the title Delivering electronic texts over the Web: The current and planned practices of the oxford text archive (Morrison, 1999). Morrison presents an overview of the Oxford Text Archive and how it integrates new forms of metadata for arts and humanities archives thorughout TEI header's marks. He uses the term digital humanities within its abstract, although not exactly as a new discipline, but lilving together with Arts and Humanities.
Picture 2. The plot shows the annual production within the Scopus database of all the documents which respond to the query "digital humanities". Three periods were selected according to detected increases in the annual production. The plot with plotly (Sievert et al., 2021).
The 20 most frequents author keywords were chosen to be by year plotted, excluding digital humanities, which was the keyword uses as query. It is quite relevant the very early appearence of the keyword metadata, rising 1999, even before than digital humanities, whisch appears for the first time in 2007 -the article Discovering interesting usage patterns in text collections: Integrating text mining with visualization (Don et al., 2007)-, keeping very long time until any other 20 most frequent keyword rises together. Looking forward the future, there are keywords that increase their appearance generally within the last 5 years, such as visualizations, cultual heritage or digital history; and other keywords that had done that more in the last 2 years, such as distant reading or machine learning.
Picture 3. To visualize better each keyword, select it by double click over the keyword in the legend. The plot with plotly (Sievert et al., 2021).
Many of the keywords have quite irregular behaviour, increasing and decreasing several times along the years, for example, linked data, collaboration or digital articles. Almost any keyword rises suddenly, and the most of the most used keywords at the end were also present at the begining.
After removing digital humanities which appears a total of 1654 times. The most frequent term after, is visualization (81 times), followed by linked data (63), cultural heritage (58), digital history (50) and collbarotaion (49). The very technical keywords such as text mining (45), semantic web (45), distant reading (44), and machine learning (42) also have relevant positions.
Picture 4. The 20 most frequent keywords after removing "digital humanities". The plot was elaborated through tidyverse (Wickham, 2021) packages and deployed with plotly (Sievert et al., 2021). .