Rcpp Wrapper for Eunjeon Project

Description

The mecab-ko and mecab-ko-dic is based on a C++ library, and POS tagging with them is useful when the spacing of source text is not correct. For integrating mecab-ko with R, Rcpp package is used for providing the basic framework.

Details

It is based on the Eunjeon Project. For Mac OSX and Linux, You need to install mecab-ko and mecab-ko-dic before install this package in R. mecab-ko: https://bitbucket.org/eunjeon/mecab-ko mecab-ko-dic: https://bitbucket.org/eunjeon/mecab-ko-dic In Windows, install_mecab(mecabLocation) function will install mecab-ko-msvc and mecab-ko-dic-msvc in user specified directory. It is operated by system command and file I/O, the speed of the analysis is slow compared to the Linux-based operating system.

Author(s)

Junhewk Kim

References

• Eunjeon project

• Wonsup Yoon, mecab-ko VC++ builds at https://github.com/Pusnow/mecab-ko-msvc, https://github.com/Pusnow/mecab-ko-dic-msvc

Examples

## Not run: 
# install.packages("devtools")
devtools::install_github("junhewk/RmecabKo")
# On Windows platform only
install_mecab("D:/Rlibs/mecab")

phrase <- # Some Korean character vectors

# For full POS tagging
pos(phrase)
# For noun extraction only
nouns(phrase)
# For tokenizing of selective morphemes
tokens_words(phrase)
# For n-grams tokenizing
tokens_ngram(phrase)

## End(Not run)

Install Mecab-Ko-Dic in Linux and Mac OSX.

Description

install_dic installs Mecab-Ko-Dic.

Usage

install_dic()

Details

This code checks and installs Mecab-Ko-Dic in Linux and Mac OSX. This is essential for using custom-defined user dictionary. Installing Mecab-Ko-Dic needs system previleges, because it uses 'make install' to build from source and install it to system.

Value

None. The function will halt when the current operation system is not Linux or Mac OSX, or Mecab-Ko-Dic is installed already.

See examples in Github.

Examples

## Not run: 
install_dic()

## End(Not run)

Install mecab-ko-msvc and mecab-ko-dic-msvc

Description

install_mecab installs Mecab-Ko-MSVC and Mecab-Ko-Dic-MSVC.

Usage

install_mecab(mecabLocation)

Arguments

Details

This code checks and installs Mecab-Ko-MSVC and Mecab-Ko-Dic-MSVC in user specified directory. Windows only.

Value

None. The function will halt when the current operation system is not Windows, or /mecabLocation/mecab.exe exists.

See examples in Github.

Examples

## Not run: 
install_mecab("D:/Rlibs/mecab")

## End(Not run)

Noun extractor by mecab-ko

Description

nouns returns nouns extracted from Korean phrases.

Usage

nouns(phrase)

Arguments

Details

Noun extraction is used for many Korean text analysis algorithms.

Value

List of nouns will be returned. Element name of the list are original phrases.

See examples in Github.

Examples

## Not run: 
nouns(c("Some Korean Phrases"))

## End(Not run)

POS tagging by mecab-ko

Description

pos returns part-of-speech (POS) tagged morpheme of Korean phrases.

Usage

pos(phrase, join = TRUE)

Arguments

Details

This is a basic function of part-of-speech tagging by mecab-ko.

Value

List of POS tagged morpheme will be returned in conjoined character vecter form. Element name of the list are original phrases. If join=FALSE, it returns list of morpheme with named with tags.

See examples in Github.

Examples

## Not run: 
pos(c("Some Korean Phrases"))
pos(c("Some Korean Phrases"), join=FALSE)

## End(Not run)

Morpheme tokenizer based on mecab-ko

Description

These tokernizer functions perform tokenization into full or selected morphemes, nouns.

Usage

token_morph(phrase, strip_punct = FALSE, strip_numeric = FALSE)

token_words(phrase, strip_punct = FALSE, strip_numeric = FALSE)

token_nouns(phrase, strip_punct = FALSE, strip_numeric = FALSE)

Arguments

Value

A list of character vectors containing the tokens, with one element in the list.

See examples in Github.

Examples

## Not run: 
txt <- # Some Korean sentence

token_morph(txt)
token_words(txt, strip_punct = FALSE)
token_nouns(txt, strip_numeric = TRUE)

## End(Not run)

N-gram tokenizer based on mecab-ko

Description

This function tokenizes inputs into n-grams. For the developmental purpose, this function offers basic n-gram (or shingle n-gram) only. Other n-gram functionality will be added later. Punctuations and numerics are stripped for this tokenizer, because in Korean n-grams those are usually useless. N-gram function is based on the selective morpheme tokenizer (token_words), but you can select other tokenizer as well.

Usage

token_ngrams(phrase, n = 3L, div = c("morph", "words", "nouns"),
  stopwords = character(), ngram_delim = " ")

Arguments

Value

A list of character vectors containing the tokens, with one element in the list.

See examples in Github.

Examples

## Not run: 
txt <- # Some Korean sentence

token_ngrams(txt)
token_ngrams(txt, n = 2)

## End(Not run)

Words extractor by mecab-ko

Description

words returns full morphemes extracted from Korean phrases.

Usage

words(phrase)

Arguments

Details

It is based on Mecab-Ko POS classification.

Value

List of full morphemes will be returned.

See examples in Github.

Examples

## Not run: 
words(c("Some Korean Phrases"))

## End(Not run)