A simple collocation-driven recognition of rhymes. Contains pre-trained models for Czech, Dutch, English, German, Russian, and Spanish poetry.
Details in P. Plecháč (2018). A Collocation-Driven Method of Discovering Rhymes (in Czech, English, and French Poetry). In Taming the Corpus: From Inflection and Lexis to Interpretation. Cham: Springer, 79-95.
! Requires eSpeak to be installed
pip install rhymetaggeror
pip3 install rhymetaggerTo annotate poems with one of the pre-trained models:
from rhymetagger import RhymeTagger
poem = [
"Tell me not, in mournful numbers,",
"Life is but an empty dream!",
"For the soul is dead that slumbers,",
"And things are not what they seem.",
"Life is real! Life is earnest!",
"And the grave is not its goal;",
"Dust thou art, to dust returnest,",
"Was not spoken of the soul.",
"Not enjoyment, and not sorrow,",
"Is our destined end or way;",
"But to act, that each tomorrow",
"Find us farther than today.",
]
rt = RhymeTagger()
rt.load_model(model='en')
rhymes = rt.tag(poem, output_format=3)
print(rhymes)
>> [1, 2, 1, 2, 3, 4, 3, 4, 5, 6, 5, 6]poem = [
"Über allen Gipfeln",
"Ist Ruh’,",
"In allen Wipfeln",
"Spürest du",
"Kaum einen Hauch;",
"Die Vögelein schweigen im Walde.",
"Warte nur, balde",
"Ruhest du auch.",
]
rt = RhymeTagger()
rt.load_model(model='de')
rhymes = rt.new_model(poem, output_format=3)
print(rhymes)
>> [1, 2, 1, 2, 3, 4, 4, 3]To train your own model:
from rhymetagger import RhymeTagger
rt = RhymeTagger()
rt.new_model(lang=ISO_CODE)
for poem in YOUR_CORPUS:
rt.add_to_model(poem)
rt.train_model()
rt.save_model(PATH_TO_FILE)| model | description |
|---|---|
| cs | Czech model (trained with Corpus of Czech Verse; 80k poems) |
| de | German model (trained with Metricalizer; 50k poems) |
| en | English model (trained with Guttenberg poetry corpus; 85k poems) |
| es | Spanish model (trained with DISCO; 9k poems) |
| nl | Dutch model (trained with Meertens Song Collection; 28k poems) |
| ru | Russian model (trained with Poetic subcorpus of Russian National Corpus; 18k poems) |
Load one of the pre-trained models or a custom model stored in JSON file
Parameters
model: string
either a name of one of the pre-trained models or path to a JSON file containing custom model
verbose:string
whether to print out info on model settings
Perform rhyme recognition
Parameters
poem: list
either a list of lines OR list of lists (stanzas > lines), each item may be either string holding text of the line OR ipa transcription (
transcribedmust beTrue) OR dict holding both orthography and ipa transcription {'text': ..., 'ipa': ...} (transcribedmust beTrue)
transcribed: boolean
whether ipa transcription is passed
output_format: int
1: returns list of indices for each line 2: returns list of indices for each rhyme 3: returns classic ABBA list where ints instead of letters
e.g. a limerick with a rhyme scheme a-a-b-b-a would be encoded as
1: [ [1,4], [0,4], [2], [3], [0,1] ] 2: [ [0,1,4], [2,3] ] 3: [ 1,1,2,2,1 ]
**kwargs
Parameters that may be used to override settings inherited from the model (
window, same_words, ngram, t_score_min, frequency_min, stanza_limit, prob_ipa_min, prob_ngram_min
Returns
rhymes: list
a list of rhymes in the requested format, see
output_format
RhymeTagger.new_model(lang, transcribed=False, window=5, syll_max=2, stress=True, vowel_length=True, ngram=1, ngram_length=3, same_words=True, t_score_min=3.078, frequency_min=3, stanza_limit=False, prob_ipa_min=0.95, prob_ngram_min = 0.95, max_iter=20, verbose=True)
Initialize new model
Parameters
lang: string
ISO language code as required by eSpeak
transcribed: boolean
whether ipa transcription is passed
window: int
how many lines forward to look for rhymes
syll_max: int
maximum number of syllables taken into account
stress: boolean
whether to focus only on sounds following after the last stress
vowel_length: boolean
whether vowel length should be taken into account
same_words: boolean
whether repetition of the same word counts as rhyme
ngram: int
upon which iteration to start taking character n-grams into account (one-based indexing, 0 = disregard n-grams completely)
ngram_length: int
length of the character n-grams
t_score_min: float
minimum value of t-score to add pair to train set
frequency_min: int
minimum number of pair occurences to add to train set
stanza_limit: boolean
whether rhymes can only appear within the same stanza
prob_ipa_min: float
minimum ipa-based probability to treat pair as rhyme
prob_ngram_min: float
minimum ngram-based probability to treat pair as rhyme
max_iter: int
maximum number of training iteratations
verbose: boolean
should progress be printed out?
Feed the model with a poem
Parameters
poem: list
either a list of lines OR list of lists (stanzas > lines), each item may be either string holding text of the line OR dict holding both orthography and ipa transcription {'text': ..., 'ipa': ...} (
transcribedmust beTrue)
Train the model fed with poems
Save the model to a JSON file
Parameters
file: string
file path