Spaces:
Sleeping
Sleeping
| import re | |
| import PyPDF2 | |
| import gradio as gr | |
| from transformers import pipeline | |
| from collections import Counter | |
| # Load NER pipeline | |
| ner_pipeline = pipeline("ner", model="dslim/bert-base-NER", tokenizer="dslim/bert-base-NER", aggregation_strategy="simple") | |
| # Load text classification model (replace with a job-role classifier if available) | |
| text_classifier = pipeline("text-classification", model="khaimait/job-title-classification-bert") | |
| def clean_resume_text(text): | |
| text = re.sub(r'http\S+', ' ', text) | |
| text = re.sub(r'#\S+', '', text) | |
| text = re.sub(r'@\S+', ' ', text) | |
| text = re.sub(r'[^\w\s]', ' ', text) | |
| text = re.sub(r'[^\x00-\x7f]', ' ', text) | |
| return re.sub(r'\s+', ' ', text).strip() | |
| def extract_resume_text(file): | |
| try: | |
| reader = PyPDF2.PdfReader(file) | |
| text = "" | |
| for page in reader.pages: | |
| page_text = page.extract_text() | |
| if page_text: | |
| text += page_text + " " | |
| if not text.strip(): | |
| return None, "Error: No text extracted from PDF." | |
| return text, None | |
| except Exception as e: | |
| return None, f"Error reading PDF: {str(e)}" | |
| def classify_resume_ner(entities): | |
| orgs = [e['word'] for e in entities if e['entity_group'] == 'ORG'] | |
| locs = [e['word'] for e in entities if e['entity_group'] == 'LOC'] | |
| jobs = [e['word'] for e in entities if e['entity_group'] == 'MISC'] | |
| dominant_org = Counter(orgs).most_common(1) | |
| dominant_loc = Counter(locs).most_common(1) | |
| dominant_job = Counter(jobs).most_common(1) | |
| return { | |
| "Main_Organization": dominant_org[0][0] if dominant_org else "Unknown", | |
| "Main_Location": dominant_loc[0][0] if dominant_loc else "Unknown", | |
| "Possible_Job/Field (NER)": dominant_job[0][0] if dominant_job else "General" | |
| } | |
| def process_resumes(files): | |
| all_results = {} | |
| for file in files: | |
| file_name = file.name.split("/")[-1] | |
| resume_text, error = extract_resume_text(file) | |
| if error: | |
| all_results[file_name] = {"error": error} | |
| continue | |
| cleaned_text = clean_resume_text(resume_text) | |
| entities = ner_pipeline(cleaned_text) | |
| classification = classify_resume_ner(entities) | |
| all_results[file_name] = { | |
| "Persons": list({e["word"] for e in entities if e["entity_group"] == "PER"}), | |
| "Organizations": list({e["word"] for e in entities if e["entity_group"] == "ORG"}), | |
| "Locations": list({e["word"] for e in entities if e["entity_group"] == "LOC"}), | |
| "Other": list({e["word"] for e in entities if e["entity_group"] not in ["PER", "ORG", "LOC"]}), | |
| "Cleaned_Text": cleaned_text, | |
| "Classification (NER)": classification | |
| } | |
| return all_results | |
| def classify_resumes_with_model(files): | |
| predictions = {} | |
| for file in files: | |
| file_name = file.name.split("/")[-1] | |
| resume_text, error = extract_resume_text(file) | |
| if error: | |
| predictions[file_name] = {"error": error} | |
| continue | |
| cleaned_text = clean_resume_text(resume_text) | |
| result = text_classifier(cleaned_text[:512]) # Truncate long resumes | |
| predictions[file_name] = { | |
| "Predicted Label (HuggingFace Classifier)": result[0]['label'], | |
| "Confidence": round(result[0]['score'], 4) | |
| } | |
| return predictions | |
| # Gradio UI | |
| with gr.Blocks(title="Multi-Resume Entity & Job Classifier") as demo: | |
| gr.Markdown("## 📂 Multi-Resume Entity Extractor & Classifier\nUpload multiple PDF resumes below. This tool extracts text, identifies key entities, and classifies job field using a Hugging Face model.") | |
| with gr.Row(): | |
| file_input = gr.File(file_types=[".pdf"], label="Upload Resume PDFs", file_count="multiple") | |
| with gr.Row(): | |
| extract_button = gr.Button("🔍 Extract & Analyze Entities") | |
| classify_button = gr.Button("🧠 Predict Job Role with Classifier") | |
| output_entities = gr.JSON(label="Entity Extraction & NER Classification") | |
| output_class = gr.JSON(label="Predicted Job Classification (Model)") | |
| extract_button.click(fn=process_resumes, inputs=[file_input], outputs=[output_entities]) | |
| classify_button.click(fn=classify_resumes_with_model, inputs=[file_input], outputs=[output_class]) | |
| if __name__ == "__main__": | |
| demo.launch() | |