Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
@@ -5,31 +5,31 @@ import spacy
|
|
5 |
import subprocess
|
6 |
import nltk
|
7 |
from nltk.corpus import wordnet
|
8 |
-
from spellchecker import SpellChecker
|
9 |
|
10 |
# Initialize the English text classification pipeline for AI detection
|
11 |
pipeline_en = pipeline(task="text-classification", model="Hello-SimpleAI/chatgpt-detector-roberta")
|
12 |
|
|
|
|
|
|
|
13 |
# Function to predict the label and score for English text (AI Detection)
|
14 |
def predict_en(text):
|
15 |
res = pipeline_en(text)[0]
|
16 |
return res['label'], res['score']
|
17 |
|
18 |
-
# Ensure necessary NLTK data is downloaded
|
19 |
nltk.download('wordnet')
|
20 |
nltk.download('omw-1.4')
|
21 |
|
22 |
-
# Ensure the SpaCy model is installed
|
23 |
try:
|
24 |
nlp = spacy.load("en_core_web_sm")
|
25 |
except OSError:
|
26 |
subprocess.run(["python", "-m", "spacy", "download", "en_core_web_sm"])
|
27 |
nlp = spacy.load("en_core_web_sm")
|
28 |
|
29 |
-
#
|
30 |
-
spell = SpellChecker()
|
31 |
-
|
32 |
-
# Function to get synonyms using NLTK WordNet (Humanifier)
|
33 |
def get_synonyms_nltk(word, pos):
|
34 |
synsets = wordnet.synsets(word, pos=pos)
|
35 |
if synsets:
|
@@ -37,7 +37,7 @@ def get_synonyms_nltk(word, pos):
|
|
37 |
return [lemma.name() for lemma in lemmas]
|
38 |
return []
|
39 |
|
40 |
-
# Function to capitalize the first letter of sentences and proper nouns
|
41 |
def capitalize_sentences_and_nouns(text):
|
42 |
doc = nlp(text)
|
43 |
corrected_text = []
|
@@ -55,7 +55,20 @@ def capitalize_sentences_and_nouns(text):
|
|
55 |
|
56 |
return ' '.join(corrected_text)
|
57 |
|
58 |
-
# Function to correct
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
59 |
def correct_singular_plural_errors(text):
|
60 |
doc = nlp(text)
|
61 |
corrected_text = []
|
@@ -77,18 +90,6 @@ def correct_singular_plural_errors(text):
|
|
77 |
|
78 |
return ' '.join(corrected_text)
|
79 |
|
80 |
-
# Function to correct tense errors in a sentence (Tense Correction)
|
81 |
-
def correct_tense_errors(text):
|
82 |
-
doc = nlp(text)
|
83 |
-
corrected_text = []
|
84 |
-
for token in doc:
|
85 |
-
if token.pos_ == "VERB" and token.dep_ in {"aux", "auxpass"}:
|
86 |
-
lemma = wordnet.morphy(token.text, wordnet.VERB) or token.text
|
87 |
-
corrected_text.append(lemma)
|
88 |
-
else:
|
89 |
-
corrected_text.append(token.text)
|
90 |
-
return ' '.join(corrected_text)
|
91 |
-
|
92 |
# Function to check and correct article errors
|
93 |
def correct_article_errors(text):
|
94 |
doc = nlp(text)
|
@@ -122,11 +123,11 @@ def replace_with_synonym(token):
|
|
122 |
|
123 |
if synonyms:
|
124 |
synonym = synonyms[0]
|
125 |
-
if token.tag_ == "VBG":
|
126 |
synonym = synonym + 'ing'
|
127 |
-
elif token.tag_ == "VBD" or token.tag_ == "VBN":
|
128 |
synonym = synonym + 'ed'
|
129 |
-
elif token.tag_ == "VBZ":
|
130 |
synonym = synonym + 's'
|
131 |
return synonym
|
132 |
return token.text
|
@@ -148,9 +149,9 @@ def ensure_subject_verb_agreement(text):
|
|
148 |
corrected_text = []
|
149 |
for token in doc:
|
150 |
if token.dep_ == "nsubj" and token.head.pos_ == "VERB":
|
151 |
-
if token.tag_ == "NN" and token.head.tag_ != "VBZ":
|
152 |
corrected_text.append(token.head.lemma_ + "s")
|
153 |
-
elif token.tag_ == "NNS" and token.head.tag_ == "VBZ":
|
154 |
corrected_text.append(token.head.lemma_)
|
155 |
corrected_text.append(token.text)
|
156 |
return ' '.join(corrected_text)
|
@@ -158,10 +159,17 @@ def ensure_subject_verb_agreement(text):
|
|
158 |
# Function to correct spelling errors
|
159 |
def correct_spelling(text):
|
160 |
words = text.split()
|
161 |
-
corrected_words = [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
162 |
return ' '.join(corrected_words)
|
163 |
|
164 |
-
# Function to paraphrase
|
165 |
def paraphrase_and_correct(text):
|
166 |
# Capitalize first to ensure proper noun capitalization
|
167 |
paraphrased_text = capitalize_sentences_and_nouns(text)
|
|
|
5 |
import subprocess
|
6 |
import nltk
|
7 |
from nltk.corpus import wordnet
|
8 |
+
from spellchecker import SpellChecker
|
9 |
|
10 |
# Initialize the English text classification pipeline for AI detection
|
11 |
pipeline_en = pipeline(task="text-classification", model="Hello-SimpleAI/chatgpt-detector-roberta")
|
12 |
|
13 |
+
# Initialize the spell checker
|
14 |
+
spell = SpellChecker()
|
15 |
+
|
16 |
# Function to predict the label and score for English text (AI Detection)
|
17 |
def predict_en(text):
|
18 |
res = pipeline_en(text)[0]
|
19 |
return res['label'], res['score']
|
20 |
|
21 |
+
# Ensure necessary NLTK data is downloaded
|
22 |
nltk.download('wordnet')
|
23 |
nltk.download('omw-1.4')
|
24 |
|
25 |
+
# Ensure the SpaCy model is installed
|
26 |
try:
|
27 |
nlp = spacy.load("en_core_web_sm")
|
28 |
except OSError:
|
29 |
subprocess.run(["python", "-m", "spacy", "download", "en_core_web_sm"])
|
30 |
nlp = spacy.load("en_core_web_sm")
|
31 |
|
32 |
+
# Function to get synonyms using NLTK WordNet
|
|
|
|
|
|
|
33 |
def get_synonyms_nltk(word, pos):
|
34 |
synsets = wordnet.synsets(word, pos=pos)
|
35 |
if synsets:
|
|
|
37 |
return [lemma.name() for lemma in lemmas]
|
38 |
return []
|
39 |
|
40 |
+
# Function to capitalize the first letter of sentences and proper nouns
|
41 |
def capitalize_sentences_and_nouns(text):
|
42 |
doc = nlp(text)
|
43 |
corrected_text = []
|
|
|
55 |
|
56 |
return ' '.join(corrected_text)
|
57 |
|
58 |
+
# Function to correct tense errors in a sentence
|
59 |
+
def correct_tense_errors(text):
|
60 |
+
doc = nlp(text)
|
61 |
+
corrected_text = []
|
62 |
+
for token in doc:
|
63 |
+
if token.pos_ == "VERB" and token.dep_ in {"aux", "auxpass"}:
|
64 |
+
# Replace with appropriate verb form
|
65 |
+
lemma = wordnet.morphy(token.text, wordnet.VERB) or token.text
|
66 |
+
corrected_text.append(lemma)
|
67 |
+
else:
|
68 |
+
corrected_text.append(token.text)
|
69 |
+
return ' '.join(corrected_text)
|
70 |
+
|
71 |
+
# Function to correct singular/plural errors
|
72 |
def correct_singular_plural_errors(text):
|
73 |
doc = nlp(text)
|
74 |
corrected_text = []
|
|
|
90 |
|
91 |
return ' '.join(corrected_text)
|
92 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
93 |
# Function to check and correct article errors
|
94 |
def correct_article_errors(text):
|
95 |
doc = nlp(text)
|
|
|
123 |
|
124 |
if synonyms:
|
125 |
synonym = synonyms[0]
|
126 |
+
if token.tag_ == "VBG": # Present participle (e.g., running)
|
127 |
synonym = synonym + 'ing'
|
128 |
+
elif token.tag_ == "VBD" or token.tag_ == "VBN": # Past tense or past participle
|
129 |
synonym = synonym + 'ed'
|
130 |
+
elif token.tag_ == "VBZ": # Third-person singular present
|
131 |
synonym = synonym + 's'
|
132 |
return synonym
|
133 |
return token.text
|
|
|
149 |
corrected_text = []
|
150 |
for token in doc:
|
151 |
if token.dep_ == "nsubj" and token.head.pos_ == "VERB":
|
152 |
+
if token.tag_ == "NN" and token.head.tag_ != "VBZ": # Singular noun, should use singular verb
|
153 |
corrected_text.append(token.head.lemma_ + "s")
|
154 |
+
elif token.tag_ == "NNS" and token.head.tag_ == "VBZ": # Plural noun, should not use singular verb
|
155 |
corrected_text.append(token.head.lemma_)
|
156 |
corrected_text.append(token.text)
|
157 |
return ' '.join(corrected_text)
|
|
|
159 |
# Function to correct spelling errors
|
160 |
def correct_spelling(text):
|
161 |
words = text.split()
|
162 |
+
corrected_words = []
|
163 |
+
for word in words:
|
164 |
+
candidates = spell.candidates(word)
|
165 |
+
if candidates:
|
166 |
+
corrected_word = spell.candidates(word).pop()
|
167 |
+
else:
|
168 |
+
corrected_word = word
|
169 |
+
corrected_words.append(corrected_word)
|
170 |
return ' '.join(corrected_words)
|
171 |
|
172 |
+
# Function to paraphrase and correct grammar
|
173 |
def paraphrase_and_correct(text):
|
174 |
# Capitalize first to ensure proper noun capitalization
|
175 |
paraphrased_text = capitalize_sentences_and_nouns(text)
|