def __init__(self, wordlist='Brown', NLength=2, prec=50):
'''
Inisiates the class, populating the frequancy data structures.
@param wordlist: This is for which master corpus to use defaultis Brown
@note: Posibal input -
Brown
AmE06
BE06
@param NLength: of Ngrams to calculate log for default 2
@param prec: the prossision of the log calculation
'''
ng = NGram()
getcontext().prec = prec
if wordlist == 'Brown':
self.fdist2 = nltk.FreqDist(nltk.corpus.brown.words())
self.fdist2Gram = nltk.FreqDist(
ng.NGramUn(nltk.corpus.brown.raw(), n=NLength)[0])
elif wordlist == 'AmE06':
self.fdist2 = nltk.FreqDist(AmE06().getCorpa().words())
self.fdist2Gram = nltk.FreqDist(
ng.NGramUn(AmE06().getCorpa().raw(), n=NLength)[0])
elif wordlist == 'BE06':
self.fdist2 = nltk.FreqDist(BE06().getCorpa().words())
self.fdist2Gram = nltk.FreqDist(
ng.NGramUn(BE06().getCorpa().raw(), n=NLength)[0])
return
def retrieve_NGram():
configs = get_configs()
if configs['corpus'] != 'rock':
fname = 'bigram-bach.pkl'
symbol_fname = 'rn2letter-bach.pkl'
else:
fname = 'bigram-rock.pkl'
symbol_fname = 'rn2letter-rock.pkl'
print '...retrieve_NGram fname', fname
ngram = NGram(fname=fname)
ngram.rn2letter, ngram.letter2rn, ngram.syms = \
update_syms(configs, symbol_fname, ngram.syms)
print 'retrieve_NGram, # of syms', len(ngram.syms)
return ngram
def gerarNGrams(self, n):
self.__lista = l.DListaEncadeada()
entrada = self.__texto.split()
lista_termos = np.empty(len(entrada) - n + 1, dtype=object)
for x in range(len(entrada) - n + 1):
lista_termos[x] = entrada[x:x + n]
for x in range(len(lista_termos)):
self.__lista.anexar(NGram(lista_termos[x], n))
def do_ngram():
print("~~~running ngram extraction~~~")
print("making samples", end=" ... ")
sample_start = process_time()
samples = []
length = 0
lmrnn = LanguageModel(rnn)
while length < args.ngram_total_sample_length:
s = lmrnn.sample(cutoff=args.ngram_max_sample_length)
samples.append(s)
length += (len(s) + 1) # ending the sequence is also a sample
ngrams = {}
ngrams_folder = rnn_folder + "/ngram"
prepare_directory(ngrams_folder)
sample_time = process_time() - sample_start
print("done, that took:", clock_str(sample_start))
print("making the actual ngrams", end=" ... ")
with open(ngrams_folder + "/samples.txt", "w") as f:
print(len(samples), len(rnn.internal_alphabet), file=f)
for s in samples:
print(len(s), *s, file=f)
for n in args.ngram_ns:
ngram_start = process_time()
ngram = NGram(n, rnn.input_alphabet, samples)
ngram.creation_info = {
"extraction time": sample_time + process_time() - ngram_start,
"size": len(ngram._state_probs_dist),
"n": n,
"total samples len (including EOS)": length,
"num samples": len(samples),
"samples cutoff len": args.ngram_max_sample_length
}
overwrite_file(ngram, ngrams_folder + "/" + str(n))
ngrams[n] = ngram
with open(ngrams_folder + "/creation_infos.txt", "w") as f:
print("ngrams made from",
len(samples),
"samples, of total length",
length,
"(including EOSs)",
file=f)
for n in ngrams:
print("===", n, "===\n", ngrams[n].creation_info, "\n\n", file=f)
print("done, that took overall", clock_str(sample_start))
return ngrams
def __init__(self,
llwl='Brown',
llNL=2,
percen=80,
NE=True,
Col=True,
Gram=True,
Chu=True):
'''
@param llwl:LogLikleyHood Corpa name ('Brown','AmE06','BE06')
@param llNL:LogLikleyHood
@param percen: Presision of output default = 20, 20% returned
@param NE: Uses NE default True
@param Col: Uses Collocation default True
@param Gram: Uses N-Grams default True
@param Chu: Uses Chunking default True
'''
self.NEs = NE
self.Col = Col
self.Gram = Gram
self.Chu = Chu
self.p = percen
print 'Starting to build ', llwl
self.LL = LogLikelihood(wordlist=llwl, NLength=llNL)
print 'LL Loaded'
self.POS = POS()
print 'POS Loaded'
self.GD = GetData()
print 'GD Loaded'
self.Cu = Chunker(self.POS)
print 'Cu Loaded'
self.FL = Filter()
print 'FL Loaded'
self.CC = Collocation(self.POS)
print 'CC Loaded'
self.Ng = NGram()
print 'Ng Loaded'
self.S = Select(percentil=self.p)
print 'S Loaded'
self.To = Tokenize(self.FL)
print 'To Loaded'
def gerarNGrams(self, n):
''' Gera os ngrams do documento no tamanho do parametro n, com base na técnica da janela deslizante'''
self.__lista = l.DListaEncadeada()
entrada = self.__texto
#busca o range até o limite da lista de palavras -n+1 para não ficar com a formatação errada nos ngram
for x in range(len(entrada)-n+1):
#recebe o indice que se refere ao ngram, ao invés de armazenar palavras
indice = (x,x+n)
#referencia do documento que gerou o ngram
doc = self
#lista duplamente encadeada que armazena todos os ngram deste documento
self.__lista.anexar(NGram(indice,doc,n))
def main():
training_set = TrainingSetHandler()
training_set.load_training_set()
gram_list = []
for size in Constants.SIZE_OF_GRAMS:
for language in training_set.language_list:
gram = NGram(size, string.ascii_lowercase, 0.5)
gram.train(training_set.training_set[language], language)
gram_list.append(gram)
#dump copies of grams to file
dump_grams(gram_list)
predic = Predictor(gram_list)
test_set_handler = TestSetHandler()
test_set_handler.load_test_sentence()
for idx, sentence in enumerate(test_set_handler.test_set):
clean_sentence = "".join([c for c in sentence[1] if c.isalpha()]).lower()
prediction = predic.predict_this_sentence(clean_sentence)
with open(os.path.join(Constants.OUTPUT_PATH, "out{}.txt".format(idx)), 'w') as f:
output = OutputHelper(prediction, sentence, f)
output.print_and_save_output()
def run_model(trainingdata, N, strings, smoothing=False):
'''
Simple helper function to run each model configuration
'''
trainingdata.seek(0)
model = NGram(N, trainingdata, smoothing)
print(model.get_model_formatted())
sys.stdout.write(os.linesep)
print("Sentence Probability Evaluation")
for s in strings:
print("Log(p) = %08.4f".ljust(20) % model.sentence_probability(s) +
"Sentence: <s> %s </s>" % s)
sys.stdout.write(os.linesep)
if smoothing:
print("Generated Sentences")
for s in model.generate_sentence(10):
print("Log(p) = %08.4f".ljust(20) % model.sentence_probability(s) +
"Sentence: %s" % s)
sys.stdout.write(os.linesep)
def get_models(data=None, configs=None, save=False):
if configs is None:
configs = get_configs()
if data is None:
data = get_data(configs)
# TODO: remove hack
if configs["bigram"]:
reduced_keys = [configs["reduced_key"]]
data.keys = reduced_keys
test_data = data.get_test_data()
test_data.keys = reduced_keys
retrieve_model = configs["retrieve_model"]
model = SkipGramNN(data, configs)
print "SkipGramNN, # of syms", len(model.syms)
if not retrieve_model:
model_loss = model.train()
if save:
model.save("skipgram-%s.pkl" % (configs["corpus"]))
plt.clf()
plt.plot(model.loss_curve)
plt.savefig("losses-%s.png" % configs.name)
print "=== train loss ==="
print "loss: %.2f" % model_loss
loss = model.check_loss()
if not configs["regularize"]:
assert np.allclose(loss, model_loss)
if save:
model_weights = model.weights.value
fname = "w1-%s.pkl" % configs.name
print fname
with open(fname, "wb") as p:
pickle.dump(model.W1.value, p)
pickle.dump(data.syms, p)
pickle.dump(model_loss, p)
pickle.dump(model_weights, p)
pickle.dump(configs, p)
fname = "skipgram-bach.pkl"
model.save(fname)
else:
fname = os.path.join("data", "test_skipgram_model.pkl")
print fname
assert fname is not None, "Error: no model to retrieve in the time being"
with open(fname, "rb") as p:
w1 = pickle.load(p)
syms = pickle.load(p)
model_loss = pickle.load(p)
model_weights = pickle.load(p)
configs_reloaded = pickle.load(p)
for key in configs.keys():
if key not in configs_reloaded.keys():
print "no key", key
for key in configs.keys():
if key in configs_reloaded.keys():
if configs[key] != configs_reloaded[key]:
print configs[key], configs_reloaded[key]
# assert configs == configs_reloaded
model.init_weights(model_weights, model_loss)
train_seq_data = data.get_train_seqs_data()
train_seqs = [seq for seq in train_seq_data.seqs]
syms = data.syms
# ngram_model = NGram(train_seqs, syms, 2, configs)
ngram_model = NGram(data.seqs, syms, 2, configs)
print "\n\nNgram, # of syms", len(ngram_model.syms)
if save:
ngram_model.save("bigram-%s.pkl" % (configs["corpus"]))
print len(ngram_model.syms), len(model.data.syms)
assert ngram_model.syms == model.data.syms
return model, ngram_model
class runable(object):
'''
Class for selecting keywords and extracting keywords from online contentent.
'''
def __init__(self,
llwl='Brown',
llNL=2,
percen=80,
NE=True,
Col=True,
Gram=True,
Chu=True):
'''
@param llwl:LogLikleyHood Corpa name ('Brown','AmE06','BE06')
@param llNL:LogLikleyHood
@param percen: Presision of output default = 20, 20% returned
@param NE: Uses NE default True
@param Col: Uses Collocation default True
@param Gram: Uses N-Grams default True
@param Chu: Uses Chunking default True
'''
self.NEs = NE
self.Col = Col
self.Gram = Gram
self.Chu = Chu
self.p = percen
print 'Starting to build ', llwl
self.LL = LogLikelihood(wordlist=llwl, NLength=llNL)
print 'LL Loaded'
self.POS = POS()
print 'POS Loaded'
self.GD = GetData()
print 'GD Loaded'
self.Cu = Chunker(self.POS)
print 'Cu Loaded'
self.FL = Filter()
print 'FL Loaded'
self.CC = Collocation(self.POS)
print 'CC Loaded'
self.Ng = NGram()
print 'Ng Loaded'
self.S = Select(percentil=self.p)
print 'S Loaded'
self.To = Tokenize(self.FL)
print 'To Loaded'
def Select(self, url, depth):
'''
Determin the best keywords for a webpage.
@param url: the base url to start sampaling from
@param depth: the depth of the website to be sampled
@return: the list of selected keywords, ordered with the highest rated words to the lower bownd of array.
'''
#Get data from web page
text = self.GD.getWebPage(url, depth)
#Tokonize sentance and words
tok = self.To.Tok(text)
#POS tag the text
pos = self.POS.POSTag(tok, 'tok')
#Log Likly Hood
log = self.LL.calcualte(tok)
#Collocations
if self.Col == True:
col = self.CC.col(pos, tok)
else:
col = []
#NE Extraction
if self.NEs == True:
ne = self.Cu.Chunks(pos,
nodes=['PERSON', 'ORGANIZATION', 'LOCATION'])
else:
ne = []
#Extract NP
if self.Chu == True:
chu = [self.Cu.parse(p) for p in pos]
else:
chu = []
#Creat N-gram
if self.Gram == True:
ga = self.Ng.Grams(pos, n=6)
else:
ga = []
return self.S.keywords(ne, ga, col, chu, log)
sys.exit()
if sys.argv[7] not in ['0', '1']:
print(
'Invalid BYOM value. Enter 1 for True or 0 for False, terminating program...'
)
sys.exit()
# Main Method
if __name__ == "__main__":
verifyArgs()
# Build and test with BYOM code,
if sys.argv[7] == '1':
byom = BYOM(sys.argv[1], float(sys.argv[3]), sys.argv[4], sys.argv[5],
bool(int(sys.argv[6])))
byom.initialize()
evalBYOM.evalBYOM(byom)
# Otherwise build and test Ngram model.
else:
ngram = NGram(sys.argv[1], int(sys.argv[2]), float(sys.argv[3]),
sys.argv[4], sys.argv[5], bool(int(sys.argv[6])))
ngram.initialize()
evaluate.evaluate(ngram)
