R package to perform sentiment analysis for Dutch/French/English and Parts of Speech tagging for Dutch/French/English/German/Spanish/Italian
The pattern.nlp package allows to do the following text mining tasks based on the pattern library
- POS tagging: Parts of Speech tagging for Dutch, French, English, German, Spanish, Italian
- Sentiment analysis: Sentiment + subjectivity scoring for Dutch, French, English
The following shows how to use the package
library(pattern.nlp)
x <- pattern_sentiment("i really really hate iphones", language = "english")
y <- pattern_sentiment("We waren bijna bij de kooien toen er van boven
een hoeragejuich losbrak alsof Rudi Vuller door Koeman in z'n kloten was geschopt.", language = "dutch")
z <- pattern_sentiment("j'aime Paris, c'est super", language = "french")
rbind(x, y, z)
polarity subjectivity id
-0.80 0.90 i really really hate iphones
0.70 1.00 We waren bijna bij de kooien ...
0.65 0.75 j'aime Paris, c'est super
x <- "Dus godvermehoeren met pus in alle puisten, zei die schele van Van Bukburg en hij had nog gelijk ook.
Er was toen dat liedje van tietenkonttieten kont tieten kontkontkont,
maar dat hoefden we geenseens niet te zingen"
pattern_pos(x = x, language = 'dutch')
x <- "Il pleure dans mon coeur comme il pleut sur la ville. Quelle est cette langueur qui penetre mon coeur?"
pattern_pos(x = x, language = 'french')
x <- "BNOSAC provides consultancy in open source analytical intelligence.
We gather dedicated open source software engineers with a focus on data mining,
business intelligence, statistical engineering and advanced artificial intelligence."
pattern_pos(x = x, language = 'english')
x <- "Der Turmer, der schaut zu Mitten der Nacht.
Hinab auf die Graber in Lage
Der Mond, der hat alles ins Helle gebracht.
Der Kirchhof, er liegt wie am Tage.
Da regt sich ein Grab und ein anderes dann."
pattern_pos(x = x, language = 'german')
x <- "Pasaron cuatro jinetes, sobre jacas andaluzas
con trajes de azul y verde, con largas capas oscuras."
pattern_pos(x = x, language = 'spanish')
x <- "Avevamo vegliato tutta la notte - i miei amici ed io sotto lampade
di moschea dalle cupole di ottone traforato, stellate come le nostre anime,
perché come queste irradiate dal chiuso fulgòre di un cuore elettrico."
pattern_pos(x = x, language = 'italian')
The output of the POS tagging shows at least the following elements:
sentence.id sentence.language chunk.id chunk.type chunk.pnp chunk.relation word.id word word.type word.lemma
9 fr 1 NP <NA> SBJ 1 Il PRP il
9 fr 2 VP <NA> <NA> 2 pleure VB pleurer
9 fr 3 PNP PNP <NA> 3 dans IN dans
9 fr 3 PNP PNP <NA> 4 mon PRP$ mon
9 fr 3 PNP PNP <NA> 5 coeur NN coeur
9 fr 4 PNP PNP <NA> 6 comme IN comme
9 fr 4 PNP PNP <NA> 7 il PRP il
9 fr 5 VP <NA> <NA> 8 pleut VB pleuvoir
9 fr 6 PNP PNP <NA> 9 sur IN sur
9 fr 6 PNP PNP <NA> 10 la DT la
9 fr 6 PNP PNP <NA> 11 ville NN ville
9 fr 7 <NA> <NA> <NA> 12 . . .
10 fr 1 NP <NA> SBJ 13 Quelle PRP quelle
10 fr 2 VP <NA> <NA> 14 est VB être
10 fr 3 NP <NA> OBJ 15 cette PRP cette
10 fr 3 NP <NA> OBJ 16 langueur NN langueur
10 fr 4 <NA> <NA> <NA> 17 qui WP qui
10 fr 5 VP <NA> <NA> 18 penetre VB penetre
10 fr 6 NP <NA> OBJ 19 mon PRP$ mon
10 fr 6 NP <NA> OBJ 20 coeur NN coeur
10 fr 7 <NA> <NA> <NA> 21 ? . ?
More information about these tags can be found at http://www.clips.ua.ac.be/pages/mbsp-tags
First install Python version 2.5+ (not version 3) and the pattern package (https://github.com/clips/pattern). Mark that the pattern package is released under the BSD license.
pip install pattern
Make sure the location of Python is into the PATH and proceed by installing the R package pattern.nlp as follows:
# Option 1: using an older version of devtools
devtools::install_github("bnosac/pattern.nlp", args = "--no-multiarch")
# Option 2: using an newer version of devtools
devtools::install_github("bnosac/pattern.nlp", INSTALL_opts = "--no-multiarch")
# Option 3: clone the repository and install the package from the command line
R CMD INSTALL pattern.nlp --no-multiarch
Make sure your when you run the R version (64/32 bit) it is the same as the Python version you installed (64/32 bit). Advise: don't use RStudio, but just plain R when executing the code. Mark that the pattern.nlp package is released under the AGPL-3 license.
If you get errors like 'No module named pattern.db' when installing the package. Note that pattern.nlp uses the following code to try to find the Python pattern package which should be installed at your computer. As long as this fails, you can not install pattern.nlp and you get the 'No module named pattern.db' error.
library(findpython)
can_find_python_cmd(required_modules = "pattern.db")
You can look at findpython::can_find_python_cmd to see which environment variables you need to set if you put Python/Pattern in a non-standard location to cover when the above error happens.
Need support in text mining. Contact BNOSAC: http://www.bnosac.be