import os import gradio as gr from transformers import pipeline import spacy import subprocess import nltk from nltk.corpus import wordnet from nltk.corpus import stopwords from nltk.tokenize import word_tokenize from spellchecker import SpellChecker import re import string import random # Download necessary NLTK data nltk.download('punkt') nltk.download('stopwords') nltk.download('averaged_perceptron_tagger') nltk.download('averaged_perceptron_tagger_eng') nltk.download('wordnet') nltk.download('omw-1.4') nltk.download('punkt_tab') # Initialize stopwords stop_words = set(stopwords.words("english")) # Words we don't want to replace exclude_tags = {'PRP', 'PRP$', 'MD', 'VBZ', 'VBP', 'VBD', 'VBG', 'VBN', 'TO', 'IN', 'DT', 'CC'} exclude_words = {'is', 'am', 'are', 'was', 'were', 'have', 'has', 'do', 'does', 'did', 'will', 'shall', 'should', 'would', 'could', 'can', 'may', 'might'} # Initialize the English text classification pipeline for AI detection pipeline_en = pipeline(task="text-classification", model="Hello-SimpleAI/chatgpt-detector-roberta") # Initialize the spell checker spell = SpellChecker() # Ensure the SpaCy model is installed try: nlp = spacy.load("en_core_web_sm") except OSError: subprocess.run(["python", "-m", "spacy", "download", "en_core_web_sm"]) nlp = spacy.load("en_core_web_sm") def plagiarism_removal(text): def plagiarism_remover(word): if word.lower() in stop_words or word.lower() in exclude_words or word in string.punctuation: return word # Find synonyms synonyms = set() for syn in wordnet.synsets(word): for lemma in syn.lemmas(): if "_" not in lemma.name() and lemma.name().isalpha() and lemma.name().lower() != word.lower(): synonyms.add(lemma.name()) pos_tag_word = nltk.pos_tag([word])[0] if pos_tag_word[1] in exclude_tags: return word filtered_synonyms = [syn for syn in synonyms if nltk.pos_tag([syn])[0][1] == pos_tag_word[1]] if not filtered_synonyms: return word synonym_choice = random.choice(filtered_synonyms) if word.istitle(): return synonym_choice.title() return synonym_choice para_split = word_tokenize(text) final_text = [plagiarism_remover(word) for word in para_split] corrected_text = [] for i in range(len(final_text)): if final_text[i] in string.punctuation and i > 0: corrected_text[-1] += final_text[i] else: corrected_text.append(final_text[i]) return " ".join(corrected_text) def predict_en(text): res = pipeline_en(text)[0] return res['label'], res['score'] def remove_redundant_words(text): doc = nlp(text) meaningless_words = {"actually", "basically", "literally", "really", "very", "just"} filtered_text = [token.text for token in doc if token.text.lower() not in meaningless_words] return ' '.join(filtered_text) def fix_punctuation_spacing(text): words = text.split(' ') cleaned_words = [] punctuation_marks = {',', '.', "'", '!', '?', ':'} for word in words: if cleaned_words and word and word[0] in punctuation_marks: cleaned_words[-1] += word else: cleaned_words.append(word) return ' '.join(cleaned_words).replace(' ,', ',').replace(' .', '.').replace(" '", "'") \ .replace(' !', '!').replace(' ?', '?').replace(' :', ':') def fix_possessives(text): text = re.sub(r'(\w)\s\'\s?s', r"\1's", text) return text def capitalize_sentences_and_nouns(text): doc = nlp(text) corrected_text = [] for sent in doc.sents: sentence = [] for token in sent: if token.i == sent.start: sentence.append(token.text.capitalize()) elif token.pos_ == "PROPN": sentence.append(token.text.capitalize()) else: sentence.append(token.text) corrected_text.append(' '.join(sentence)) return ' '.join(corrected_text) def force_first_letter_capital(text): sentences = re.split(r'(?<=\w[.!?])\s+', text) capitalized_sentences = [] for sentence in sentences: if sentence: capitalized_sentence = sentence[0].capitalize() + sentence[1:] if not re.search(r'[.!?]$', capitalized_sentence): capitalized_sentence += '.' capitalized_sentences.append(capitalized_sentence) return " ".join(capitalized_sentences) def correct_tense_errors(text): doc = nlp(text) corrected_text = [] for token in doc: if token.pos_ == "VERB" and token.dep_ in {"aux", "auxpass"}: lemma = wordnet.morphy(token.text, wordnet.VERB) or token.text corrected_text.append(lemma) else: corrected_text.append(token.text) return ' '.join(corrected_text) def correct_article_errors(text): doc = nlp(text) corrected_text = [] for token in doc: if token.text in ['a', 'an']: next_token = token.nbor(1) if token.text == "a" and next_token.text[0].lower() in "aeiou": corrected_text.append("an") elif token.text == "an" and next_token.text[0].lower() not in "aeiou": corrected_text.append("a") else: corrected_text.append(token.text) else: corrected_text.append(token.text) return ' '.join(corrected_text) def ensure_subject_verb_agreement(text): doc = nlp(text) corrected_text = [] for token in doc: if token.dep_ == "nsubj" and token.head.pos_ == "VERB": if token.tag_ == "NN" and token.head.tag_ != "VBZ": corrected_text.append(token.head.lemma_ + "s") elif token.tag_ == "NNS" and token.head.tag_ == "VBZ": corrected_text.append(token.head.lemma_) corrected_text.append(token.text) return ' '.join(corrected_text) def correct_spelling(text): words = word_tokenize(text) corrected_words = [] for word in words: corrected_word = spell.candidates(word) if corrected_word: corrected_words.append(spell.candidates(word).pop()) # Choose the first candidate as the correction else: corrected_words.append(word) # If it's not misspelled, keep the original word return ' '.join(corrected_words) def paraphrase_and_correct(text): paragraphs = text.split("\n\n") # Split by paragraphs # Process each paragraph separately processed_paragraphs = [] for paragraph in paragraphs: cleaned_text = remove_redundant_words(paragraph) plag_removed = plagiarism_removal(cleaned_text) paraphrased_text = capitalize_sentences_and_nouns(plag_removed) paraphrased_text = force_first_letter_capital(paraphrased_text) paraphrased_text = correct_article_errors(paraphrased_text) paraphrased_text = correct_tense_errors(paraphrased_text) paraphrased_text = ensure_subject_verb_agreement(paraphrased_text) paraphrased_text = fix_possessives(paraphrased_text) paraphrased_text = correct_spelling(paraphrased_text) # Spelling correction paraphrased_text = fix_punctuation_spacing(paraphrased_text) processed_paragraphs.append(paraphrased_text) return "\n\n".join(processed_paragraphs) # Reassemble the text with paragraphs # Gradio app setup with gr.Blocks() as demo: with gr.Tab("AI Detection"): t1 = gr.Textbox(lines=5, label='Text') button1 = gr.Button("🤖 Predict!") label1 = gr.Textbox(lines=1, label='Predicted Label 🎃') score1 = gr.Textbox(lines=1, label='Prob') button1.click(fn=predict_en, inputs=t1, outputs=[label1, score1]) with gr.Tab("Paraphrasing & Grammar Correction"): t2 = gr.Textbox(lines=5, label='Enter text for paraphrasing and grammar correction') button2 = gr.Button("🔄 Paraphrase and Correct") result2 = gr.Textbox(lines=5, label='Corrected Text') button2.click(fn=paraphrase_and_correct, inputs=t2, outputs=result2) demo.launch(share=True)