A slow version of the Rapid Automatic Keyword Extraction (RAKE) algorithm
You can get the stable version from CRAN:
install.packages("slowraker")
Or the development version from GitHub:
if (!require(devtools)) install.packages("devtools")
devtools::install_github("crew102/slowraker")
There is one main function in the slowraker
package - slowrake()
. slowrake()
extracts keywords from a vector of documents using the RAKE algorithm. This algorithm doesn’t require any training data, so it’s super easy to use:
library(slowraker)
data("dog_pubs")
rakelist <- slowrake(txt = dog_pubs$abstract[1:5])
#> Warning: package 'slowraker' was built under R version 3.4.2
slowrake()
outputs a list of data frames. Each data frame contains the keywords that were extracted for an element of txt
:
rakelist
#>
#> # A rakelist containing 5 data frames:
#> $ :'data.frame': 61 obs. of 4 variables:
#> ..$ keyword:"assistance dog identification tags" ...
#> ..$ freq :1 1 ...
#> ..$ score :11 ...
#> ..$ stem :"assist dog identif tag" ...
#> $ :'data.frame': 90 obs. of 4 variables:
#> ..$ keyword:"current dog suitability assessments focus" ...
#> ..$ freq :1 1 ...
#> ..$ score :21 ...
#> ..$ stem :"current dog suitabl assess focu" ...
#> #...With 3 more data frames.
You can bind these data frames together using rbind_rakelist()
:
rakedf <- rbind_rakelist(rakelist = rakelist, doc_id = dog_pubs$doi[1:5])
head(rakedf, 5)
#> doc_id keyword freq score
#> 1 10.1371/journal.pone.0132820 assistance dog identification tags 1 10.8
#> 2 10.1371/journal.pone.0132820 animal control facilities 1 9.0
#> 3 10.1371/journal.pone.0132820 emotional support animals 1 9.0
#> 4 10.1371/journal.pone.0132820 small body sizes 1 9.0
#> 5 10.1371/journal.pone.0132820 seemingly inappropriate dogs 1 7.9
#> stem
#> 1 assist dog identif tag
#> 2 anim control facil
#> 3 emot support anim
#> 4 small bodi size
#> 5 seemingli inappropri dog
slowrake()
, check out the “Getting started” vignette (vignette("getting-started")
). Frequently asked questions are answered in the FAQs vignette (vignette("faqs")
).