rake: Rapid automatic extraction of keywords from documents

In lmkirvan/rakeR: An implementation of rapid automatic keyword extraction for R.

Description

This is the main functions for extracting and returning ranked keywords. This algorithm works by tokenizing words and phrases based on stopwords. After tokenization the frequency of tokens and their degree \(i.e. their coocurrence with words in the remaining phrases\) are used to rank phrases. Stopwords can include both common english words and can be enhanced by including a vector of domain specific stop words.

Usage

rake(x, split_words = smart_stop_words(), split_punct = basic_punct(),
  top_fraction = 1/3)

Arguments

x	a character vector of texts to find keywords for
top_fraction	the fraction of the most highly ranked phrases to return
method	one of "degreeFreq", "degree", or "freq". degreeFreq is simply degree // frequency.

Value

returns a list with elements composed of one named integer vector for each document. key phrases are names and values are ranked

Examples

rake(test_text, 5, "degree")
rake(test_text, 15)

lmkirvan/rakeR documentation built on May 14, 2019, 1:46 p.m.