mgaughan/24_deb_pkg_gov
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

first draft of the topic model

Browse Source
This commit is contained in:
Matthew Gaughan 2024-05-02 10:55:50 -05:00
parent 793ddcc632
commit 9d9bb8da1e
1 changed files with 13 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

17
text_analysis/topicModel.py
View File

@ -26,6 +26,7 @@ from nltk.corpus import stopwords
from nltk.stem.wordnet import WordNetLemmatizer
stopwords = stopwords.words('english')
#https://nlp.stanford.edu/IR-book/html/htmledition/dropping-common-terms-stop-words-1.html
#loading data in, getting misc descriptors
def get_data_from_dir(directory):
@ -45,9 +46,13 @@ def get_data_from_dir(directory):
#preprocessing text data
def preprocess(corpus_list):
#extending stopwords
specific_stopwords = ["http", "com", "www", "org", "file", "code", "time", "software", "use", "user", "set", "line", "run", "source", "github",
"lineno", "python", "php", "ruby", "api"]
stopwords.extend(specific_stopwords)
D = copy.copy(corpus_list)
#mvp right now, can certainly be expanded as iterations of text analysis are done
D = [[token for token in simple_preprocess(doc) if token not in stopwords]for doc in D]
D = [[token for token in simple_preprocess(doc) if token not in stopwords and len(token) > 2]for doc in D]
lemmatizer = WordNetLemmatizer()
D_lemma = [" ".join([lemmatizer.lemmatize(token) for token in doc]) for doc in D]
return D_lemma
@ -63,7 +68,7 @@ def text_preparation(lemmatized_text):
D_bigrams[i].append(token)
#id2word
id2word = corpora.Dictionary(D_bigrams)
id2word.filter_extremes(no_below=2, no_above=0.5)
id2word.filter_extremes(no_below=5, no_above=0.5)
#bow representation
bag_of_words = [id2word.doc2bow(doc) for doc in D_bigrams]
return bag_of_words, id2word
@ -80,7 +85,7 @@ def lda_model_identification(data_vectorized):
print("Model Perplexity: ", best_lda_model.perplexity(data_vectorized))
#TODO: implement best LDA model here
def best_lda_model(data_vectorized):
def best_lda_model(data_vectorized, vocab):
#Best Log Likelihood Score: -502085.9749390023
#Model Perplexity: 1689.0943431883845
lda = LatentDirichletAllocation(n_components=5, learning_decay = 0.9, batch_size = 128, max_iter = 50)
@ -92,6 +97,7 @@ def best_lda_model(data_vectorized):
for topic, words in topic_words.items():
print('Topic: %d' % topic)
print(' %s' % ', '.join(words))
#lda.print_topics(num_words=10)
#TODO: evaluate model and identified topics
@ -114,6 +120,9 @@ if __name__ == "__main__":
)
data_vectorized = vectorizer.fit_transform(lemmatized_corpus)
#lda_model_identification(data_vectorized)
best_lda_model(data_vectorized)
#freqs = zip(vectorizer.get_feature_names_out(), data_vectorized.sum(axis=0).tolist()[0])
# sort from largest to smallest
#print(sorted(freqs, key=lambda x: -x[1])[:25])
best_lda_model(data_vectorized, vectorizer.get_feature_names_out())