I keep seeing nice little wordclouds summarizing research profiles from different groups. I’ve played with them myself. Drawing is simple, but getting the data can be a pain. Now, with the new
rorcid package that is made easy. If you don’t already have them, you need the following installed:
## install packages using the following install.packages("devtools") # to install from github library(devtools) install_github("ropensci/rorcid") # to access orcid through R install.packages(c("tm","wordcloud")) # to draw word clouds
I’ve put a few convenience functions in a gist, so do
f <- tempfile() download.file("https://gist.githubusercontent.com/chr1swallace/d4b54bf273006ea0acc1/raw/54ddf5065a43fabba8f0414366d1853dd6b6608d/orcidcloud.R",f,method="wget") source(f)
Then head off to ORCiD and find the id for your favourite scientist (I chose Sylvia Richardson, a Bayesian statistician and Director of the MRC Biostatistics Unit), and draw a wordcloud from their papers:
id <- "0000-0003-1998-492X" data <- orcid_id(orcid = id, profile="works") ## extract items years <- get.years(data[]) titles <- get.titles(data[]) authors <- get.authors(data[],surnames=TRUE) co.authors <- lapply(authors,setdiff,"Richardson") ## make a word cloud make.cloud(titles)
Alternatively, the whole thing can be run at once (what can be “titles”, “years”, “authors”).
data <- orcid_id(orcid = id, profile="works") orcid.cloud(data, what="titles")
You can play with drawing co authors similarly. I’ve tried also to group by decade, to see the changing research profile, but it needs a bit of work, I think. Grouping by decade, the one paper from the 1970s gets over prominant. For now, I dropped it, but there should be a better way of dealing with this.
use <- which(years >= 1980) make.cloud(titles[use],by=years[use],group=10,min.freq=2)