I wrote ridiculous keyword articles and product descriptions that aren't even close to best practices today.
Moreover, a new method named "word cluster analysis" (Mao et al. 2010), which combined the author keywords, article title, and 'KeyWords Plus', had been successfully applied to trace the hotspot of related research field in previous work (Tanaka and Ho 2011; Huang et al. 2012; Fu et al. 2013; Tan et al. 2014).
This involved the development of search strategies and keywords, article selection processes, quality assessment tools, and guidelines for the synthesis of results.
The technique of statistical analysis of author keywords, articles title, article abstract, and keyword plus could be used to identify directions in science, and has proved to be efficient in comprehending and grasping the progress of new frontiers of science.
'KeyWords Plus', which were different from the author keywords, articles title and articles abstract, were automatically generated by the SCI from an algorithm that extracts key words from all the titles referenced or cited in the documents (Garfield 1990).
However, "water and wastewater treatment", "electrooxidation or electrochemical oxidation", and "electrode(s) or titanium dioxide TIO2)" were present in words in the author keywords, articles title, and 'KeyWords Plus' list, which still might be identified as current electrochemical technology for water and wastewater treatment research hotspots.
Sadly, the search engines are not fooled anymore and easily sidestep such dense-keyword articles.
While I know that it's difficult, in the beginning at least, to write a highly-targeted-and-keyword-focused article with a minimum keyword density, whatever you write still has to sound natural and not forced.
Using a keyword based article selection has the risk of generating a rather narrow or biased dataset, especially when the aim is named entity recognition.
