Create app.py
Browse files
app.py
ADDED
|
@@ -0,0 +1,151 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gr
|
| 2 |
+
import nltk
|
| 3 |
+
from nltk.tokenize import sent_tokenize, word_tokenize
|
| 4 |
+
from nltk.corpus import stopwords, wordnet, brown
|
| 5 |
+
from nltk.stem import PorterStemmer, WordNetLemmatizer
|
| 6 |
+
from nltk import pos_tag, ne_chunk, ngrams
|
| 7 |
+
from nltk.collocations import BigramCollocationFinder
|
| 8 |
+
from nltk.classify import NaiveBayesClassifier
|
| 9 |
+
from nltk.corpus import movie_reviews
|
| 10 |
+
import random
|
| 11 |
+
|
| 12 |
+
# Tải các tài nguyên cần thiết
|
| 13 |
+
nltk.download('punkt')
|
| 14 |
+
nltk.download('punkt_tab')
|
| 15 |
+
nltk.download('averaged_perceptron_tagger')
|
| 16 |
+
nltk.download('maxent_ne_chunker')
|
| 17 |
+
nltk.download('words')
|
| 18 |
+
nltk.download('stopwords')
|
| 19 |
+
nltk.download('wordnet')
|
| 20 |
+
nltk.download('brown')
|
| 21 |
+
nltk.download('movie_reviews')
|
| 22 |
+
|
| 23 |
+
# Khởi tạo các công cụ
|
| 24 |
+
stemmer = PorterStemmer()
|
| 25 |
+
lemmatizer = WordNetLemmatizer()
|
| 26 |
+
stop_words = set(stopwords.words('english'))
|
| 27 |
+
|
| 28 |
+
# Hàm huấn luyện classifier đơn giản
|
| 29 |
+
def train_classifier():
|
| 30 |
+
pos_reviews = [({"word": word}, 'positive') for word in movie_reviews.words('pos')[:100]]
|
| 31 |
+
neg_reviews = [({"word": word}, 'negative') for word in movie_reviews.words('neg')[:100]]
|
| 32 |
+
train_set = pos_reviews + neg_reviews
|
| 33 |
+
random.shuffle(train_set)
|
| 34 |
+
return NaiveBayesClassifier.train(train_set)
|
| 35 |
+
|
| 36 |
+
classifier = train_classifier()
|
| 37 |
+
|
| 38 |
+
# Hàm chính xử lý các chức năng
|
| 39 |
+
def nlp_tool(input_text, function):
|
| 40 |
+
if not input_text:
|
| 41 |
+
return "Vui lòng nhập văn bản!"
|
| 42 |
+
|
| 43 |
+
if function == "Sentence Tokenization":
|
| 44 |
+
return "\n".join(sent_tokenize(input_text))
|
| 45 |
+
|
| 46 |
+
elif function == "Word Tokenization":
|
| 47 |
+
return "\n".join(word_tokenize(input_text))
|
| 48 |
+
|
| 49 |
+
elif function == "Part-of-Speech Tagging":
|
| 50 |
+
words = word_tokenize(input_text)
|
| 51 |
+
return "\n".join([f"{word}: {tag}" for word, tag in pos_tag(words)])
|
| 52 |
+
|
| 53 |
+
elif function == "Stemming":
|
| 54 |
+
words = word_tokenize(input_text)
|
| 55 |
+
return "\n".join([stemmer.stem(word) for word in words])
|
| 56 |
+
|
| 57 |
+
elif function == "Lemmatization":
|
| 58 |
+
words = word_tokenize(input_text)
|
| 59 |
+
return "\n".join([lemmatizer.lemmatize(word) for word in words])
|
| 60 |
+
|
| 61 |
+
elif function == "Remove Stop Words":
|
| 62 |
+
words = word_tokenize(input_text)
|
| 63 |
+
return "\n".join([word for word in words if word.lower() not in stop_words])
|
| 64 |
+
|
| 65 |
+
elif function == "Named Entity Recognition":
|
| 66 |
+
words = word_tokenize(input_text)
|
| 67 |
+
pos_tags = pos_tag(words)
|
| 68 |
+
entities = ne_chunk(pos_tags)
|
| 69 |
+
return str(entities)
|
| 70 |
+
|
| 71 |
+
elif function == "Text Classification":
|
| 72 |
+
words = word_tokenize(input_text)
|
| 73 |
+
result = classifier.classify({word: True for word in words})
|
| 74 |
+
return f"Sentiment: {result}"
|
| 75 |
+
|
| 76 |
+
elif function == "N-grams (Bigrams)":
|
| 77 |
+
words = word_tokenize(input_text)
|
| 78 |
+
bigrams = list(ngrams(words, 2))
|
| 79 |
+
return "\n".join([f"{w1} - {w2}" for w1, w2 in bigrams])
|
| 80 |
+
|
| 81 |
+
elif function == "Collocations":
|
| 82 |
+
words = word_tokenize(input_text)
|
| 83 |
+
finder = BigramCollocationFinder.from_words(words)
|
| 84 |
+
collocations = finder.nbest(nltk.collocations.BigramAssocMeasures().pmi, 5)
|
| 85 |
+
return "\n".join([f"{w1} - {w2}" for w1, w2 in collocations])
|
| 86 |
+
|
| 87 |
+
elif function == "WordNet Synsets":
|
| 88 |
+
words = word_tokenize(input_text)
|
| 89 |
+
first_word = words[0] if words else ""
|
| 90 |
+
synsets = wordnet.synsets(first_word)
|
| 91 |
+
if synsets:
|
| 92 |
+
return f"Definition: {synsets[0].definition()}\nExamples: {synsets[0].examples()}"
|
| 93 |
+
return "Không tìm thấy từ trong WordNet!"
|
| 94 |
+
|
| 95 |
+
elif function == "Sample from Brown Corpus":
|
| 96 |
+
return " ".join(brown.words()[:50])
|
| 97 |
+
|
| 98 |
+
return "Chức năng chưa được triển khai!"
|
| 99 |
+
|
| 100 |
+
# Tạo giao diện Gradio
|
| 101 |
+
with gr.Blocks(theme=gr.themes.Soft()) as demo:
|
| 102 |
+
gr.Markdown(
|
| 103 |
+
"""
|
| 104 |
+
# Công cụ xử lý ngôn ngữ tự nhiên với NLTK
|
| 105 |
+
Nhập văn bản và chọn chức năng để khám phá các khả năng của NLTK!
|
| 106 |
+
"""
|
| 107 |
+
)
|
| 108 |
+
|
| 109 |
+
with gr.Row():
|
| 110 |
+
with gr.Column(scale=1):
|
| 111 |
+
input_text = gr.Textbox(
|
| 112 |
+
label="Nhập văn bản",
|
| 113 |
+
placeholder="Ví dụ: I love coding in Python.",
|
| 114 |
+
lines=5
|
| 115 |
+
)
|
| 116 |
+
function_dropdown = gr.Dropdown(
|
| 117 |
+
label="Chọn chức năng",
|
| 118 |
+
choices=[
|
| 119 |
+
"Sentence Tokenization",
|
| 120 |
+
"Word Tokenization",
|
| 121 |
+
"Part-of-Speech Tagging",
|
| 122 |
+
"Stemming",
|
| 123 |
+
"Lemmatization",
|
| 124 |
+
"Remove Stop Words",
|
| 125 |
+
"Named Entity Recognition",
|
| 126 |
+
"Text Classification",
|
| 127 |
+
"N-grams (Bigrams)",
|
| 128 |
+
"Collocations",
|
| 129 |
+
"WordNet Synsets",
|
| 130 |
+
"Sample from Brown Corpus"
|
| 131 |
+
],
|
| 132 |
+
value="Sentence Tokenization"
|
| 133 |
+
)
|
| 134 |
+
submit_btn = gr.Button("Xử lý", variant="primary")
|
| 135 |
+
|
| 136 |
+
with gr.Column(scale=2):
|
| 137 |
+
output_text = gr.Textbox(
|
| 138 |
+
label="Kết quả",
|
| 139 |
+
lines=10,
|
| 140 |
+
interactive=False
|
| 141 |
+
)
|
| 142 |
+
|
| 143 |
+
# Kết nối nút bấm với hàm xử lý
|
| 144 |
+
submit_btn.click(
|
| 145 |
+
fn=nlp_tool,
|
| 146 |
+
inputs=[input_text, function_dropdown],
|
| 147 |
+
outputs=output_text
|
| 148 |
+
)
|
| 149 |
+
|
| 150 |
+
# Chạy giao diện
|
| 151 |
+
demo.launch()
|