Spaces:
Sleeping
Sleeping
import gradio as gr | |
import pandas as pd | |
from sentence_transformers import SentenceTransformer | |
from sklearn.metrics.pairwise import cosine_similarity | |
import csv | |
import io | |
import tempfile | |
import os | |
# νκ΅μ΄ μ²λ¦¬λ₯Ό μν KoSentence-BERT λͺ¨λΈ λ‘λ | |
model = SentenceTransformer('jhgan/ko-sbert-sts') | |
# μ μ λ³μ | |
global_recommendations = None | |
global_csv_file = None | |
youtube_columns = None | |
# CSV νμΌ μμ± ν¨μ | |
def create_csv_file(recommendations): | |
with tempfile.NamedTemporaryFile(mode='w+', delete=False, suffix='.csv', encoding='utf-8') as temp_file: | |
writer = csv.writer(temp_file) | |
writer.writerow(["Employee ID", "Employee Name", "Recommended Programs", "Recommended YouTube Content"]) | |
for rec in recommendations: | |
writer.writerow(rec) | |
return temp_file.name | |
# μ΄ λ§€μΉ ν¨μ | |
def auto_match_columns(df, required_cols): | |
matched_cols = {} | |
for req_col in required_cols: | |
matched_col = None | |
for col in df.columns: | |
if req_col.lower() in col.lower(): | |
matched_col = col | |
break | |
matched_cols[req_col] = matched_col | |
return matched_cols | |
# μ΄ κ²μ¦ ν¨μ | |
def validate_and_get_columns(employee_df, program_df): | |
required_employee_cols = ["employee_id", "employee_name", "current_skills"] | |
required_program_cols = ["program_name", "skills_acquired", "duration"] | |
employee_cols = auto_match_columns(employee_df, required_employee_cols) | |
program_cols = auto_match_columns(program_df, required_program_cols) | |
for key, value in employee_cols.items(): | |
if value is None: | |
return f"μ§μ λ°μ΄ν°μμ '{key}' μ΄μ μ νν μ μμ΅λλ€. μ¬λ°λ₯Έ μ΄μ μ ννμΈμ.", None, None | |
for key, value in program_cols.items(): | |
if value is None: | |
return f"νλ‘κ·Έλ¨ λ°μ΄ν°μμ '{key}' μ΄μ μ νν μ μμ΅λλ€. μ¬λ°λ₯Έ μ΄μ μ ννμΈμ.", None, None | |
return None, employee_cols, program_cols | |
# μ νλΈ λ°μ΄ν° μ΄ μ ν ν¨μ | |
def select_youtube_columns(youtube_file): | |
global youtube_columns | |
if youtube_file is None: | |
return [gr.Dropdown(choices=[], value="") for _ in range(4)] | |
youtube_df = pd.read_csv(youtube_file.name) | |
required_youtube_cols = ["title", "description", "url", "upload_date"] | |
youtube_columns = auto_match_columns(youtube_df, required_youtube_cols) | |
column_options = youtube_df.columns.tolist() | |
return [ | |
gr.Dropdown(choices=column_options, value=youtube_columns.get("title", "")), | |
gr.Dropdown(choices=column_options, value=youtube_columns.get("description", "")), | |
gr.Dropdown(choices=column_options, value=youtube_columns.get("url", "")), | |
gr.Dropdown(choices=column_options, value=youtube_columns.get("upload_date", "")) | |
] | |
# μ νλΈ μ½ν μΈ λ°μ΄ν° λ‘λ λ° μ²λ¦¬ ν¨μ | |
def load_youtube_content(file_path, title_col, description_col, url_col, upload_date_col): | |
youtube_df = pd.read_csv(file_path) | |
selected_columns = [col for col in [title_col, description_col, url_col, upload_date_col] if col] | |
youtube_df = youtube_df[selected_columns] | |
column_mapping = { | |
title_col: 'title', | |
description_col: 'description', | |
url_col: 'url', | |
upload_date_col: 'upload_date' | |
} | |
youtube_df.rename(columns=column_mapping, inplace=True) | |
if 'upload_date' in youtube_df.columns: | |
youtube_df['upload_date'] = pd.to_datetime(youtube_df['upload_date'], errors='coerce') | |
return youtube_df | |
# μ νλΈ μ½ν μΈ μ κ΅μ‘ νλ‘κ·Έλ¨ λ§€μΉ ν¨μ | |
def match_youtube_content(program_skills, youtube_df, model): | |
if 'description' not in youtube_df.columns: | |
return None | |
youtube_embeddings = model.encode(youtube_df['description'].tolist()) | |
program_embeddings = model.encode(program_skills) | |
similarities = cosine_similarity(program_embeddings, youtube_embeddings) | |
return similarities | |
# μ§μ λ°μ΄ν°λ₯Ό λΆμνμ¬ κ΅μ‘ νλ‘κ·Έλ¨μ μΆμ²νκ³ , ν μ΄λΈμ μμ±νλ ν¨μ | |
def hybrid_rag(employee_file, program_file, youtube_file, title_col, description_col, url_col, upload_date_col): | |
global global_recommendations | |
global global_csv_file | |
# μ§μ λ° νλ‘κ·Έλ¨ λ°μ΄ν° λ‘λ | |
employee_df = pd.read_csv(employee_file.name) | |
program_df = pd.read_csv(program_file.name) | |
error_msg, employee_cols, program_cols = validate_and_get_columns(employee_df, program_df) | |
if error_msg: | |
return error_msg, None, None | |
employee_skills = employee_df[employee_cols["current_skills"]].tolist() | |
program_skills = program_df[program_cols["skills_acquired"]].tolist() | |
employee_embeddings = model.encode(employee_skills) | |
program_embeddings = model.encode(program_skills) | |
similarities = cosine_similarity(employee_embeddings, program_embeddings) | |
# μ νλΈ μ½ν μΈ λ‘λ λ° μ²λ¦¬ | |
youtube_df = load_youtube_content(youtube_file.name, title_col, description_col, url_col, upload_date_col) | |
# μ νλΈ μ½ν μΈ μ κ΅μ‘ νλ‘κ·Έλ¨ λ§€μΉ | |
youtube_similarities = match_youtube_content(program_df[program_cols['skills_acquired']].tolist(), youtube_df, model) | |
recommendations = [] | |
recommendation_rows = [] | |
for i, employee in employee_df.iterrows(): | |
recommended_programs = [] | |
recommended_youtube = [] | |
for j, program in program_df.iterrows(): | |
if similarities[i][j] > 0.5: | |
recommended_programs.append(f"{program[program_cols['program_name']]} ({program[program_cols['duration']]})") | |
if youtube_similarities is not None: | |
top_youtube_indices = youtube_similarities[j].argsort()[-3:][::-1] # μμ 3κ° | |
for idx in top_youtube_indices: | |
if 'title' in youtube_df.columns and 'url' in youtube_df.columns: | |
recommended_youtube.append(f"{youtube_df.iloc[idx]['title']} (URL: {youtube_df.iloc[idx]['url']})") | |
# μΆμ² νλ‘κ·Έλ¨ λ° μ νλΈ μ½ν μΈ κ°μ μ ν | |
recommended_programs = recommended_programs[:5] # μ΅λ 5κ° νλ‘κ·Έλ¨λ§ μΆμ² | |
recommended_youtube = recommended_youtube[:3] # μ΅λ 3κ° μ νλΈ μ½ν μΈ λ§ μΆμ² | |
if recommended_programs: | |
recommendation = f"μ§μ {employee[employee_cols['employee_name']]}μ μΆμ² νλ‘κ·Έλ¨: {', '.join(recommended_programs)}" | |
youtube_recommendation = f"μΆμ² μ νλΈ μ½ν μΈ : {', '.join(recommended_youtube)}" if recommended_youtube else "μΆμ²ν μ νλΈ μ½ν μΈ κ° μμ΅λλ€." | |
recommendation_rows.append([employee[employee_cols['employee_id']], employee[employee_cols['employee_name']], | |
", ".join(recommended_programs), ", ".join(recommended_youtube)]) | |
else: | |
recommendation = f"μ§μ {employee[employee_cols['employee_name']]}μκ² μ ν©ν νλ‘κ·Έλ¨μ΄ μμ΅λλ€." | |
youtube_recommendation = "μΆμ²ν μ νλΈ μ½ν μΈ κ° μμ΅λλ€." | |
recommendation_rows.append([employee[employee_cols['employee_id']], employee[employee_cols['employee_name']], | |
"μ ν©ν νλ‘κ·Έλ¨ μμ", "μΆμ² μ½ν μΈ μμ"]) | |
recommendations.append(recommendation + "\n" + youtube_recommendation) | |
global_recommendations = recommendation_rows | |
# CSV νμΌ μμ± | |
global_csv_file = create_csv_file(recommendation_rows) | |
# κ²°κ³Ό ν μ΄λΈ λ°μ΄ν°νλ μ μμ± | |
result_df = pd.DataFrame(recommendation_rows, columns=["Employee ID", "Employee Name", "Recommended Programs", "Recommended YouTube Content"]) | |
return result_df, gr.File(value=global_csv_file, visible=True), gr.Button(value="CSV λ€μ΄λ‘λ", visible=True) | |
# μ±ν μλ΅ ν¨μ | |
def chat_response(message, history): | |
global global_recommendations | |
if global_recommendations is None: | |
return "λ¨Όμ 'λΆμ μμ' λ²νΌμ λλ¬ λ°μ΄ν°λ₯Ό λΆμν΄μ£ΌμΈμ." | |
for employee in global_recommendations: | |
if employee[1].lower() in message.lower(): | |
return f"{employee[1]}λμκ² μΆμ²λ νλ‘κ·Έλ¨μ λ€μκ³Ό κ°μ΅λλ€: {employee[2]}\n\nμΆμ² μ νλΈ μ½ν μΈ : {employee[3]}" | |
return "μ£μ‘ν©λλ€. ν΄λΉ μ§μμ μ 보λ₯Ό μ°Ύμ μ μμ΅λλ€. λ€λ₯Έ μ§μ μ΄λ¦μ μ λ ₯ν΄μ£ΌμΈμ." | |
# CSV λ€μ΄λ‘λ ν¨μ | |
def download_csv(): | |
global global_csv_file | |
return gr.File(value=global_csv_file, visible=True) | |
# Gradio λΈλ‘ | |
with gr.Blocks(css=".gradio-button {background-color: #007bff; color: white;} .gradio-textbox {border-color: #6c757d;}") as demo: | |
gr.Markdown("<h1 style='text-align: center; color: #2c3e50;'>πΌ HybridRAG μμ€ν (μ νλΈ μ½ν μΈ ν¬ν¨)</h1>") | |
with gr.Row(): | |
with gr.Column(scale=1, min_width=300): | |
gr.Markdown("<h3 style='color: #34495e;'>1. λ°μ΄ν°λ₯Ό μ λ‘λνμΈμ</h3>") | |
employee_file = gr.File(label="μ§μ λ°μ΄ν° μ λ‘λ", interactive=True) | |
program_file = gr.File(label="κ΅μ‘ νλ‘κ·Έλ¨ λ°μ΄ν° μ λ‘λ", interactive=True) | |
youtube_file = gr.File(label="μ νλΈ μ½ν μΈ λ°μ΄ν° μ λ‘λ", interactive=True) | |
gr.Markdown("<h4 style='color: #34495e;'>μ νλΈ λ°μ΄ν° μ΄ μ ν</h4>") | |
title_col = gr.Dropdown(label="μ λͺ© μ΄") | |
description_col = gr.Dropdown(label="μ€λͺ μ΄") | |
url_col = gr.Dropdown(label="URL μ΄") | |
upload_date_col = gr.Dropdown(label="μ λ‘λ λ μ§ μ΄") | |
youtube_file.change(select_youtube_columns, inputs=[youtube_file], outputs=[title_col, description_col, url_col, upload_date_col]) | |
analyze_button = gr.Button("λΆμ μμ", elem_classes="gradio-button") | |
output_table = gr.DataFrame(label="λΆμ κ²°κ³Ό (ν μ΄λΈ)") | |
csv_download = gr.File(label="μΆμ² κ²°κ³Ό λ€μ΄λ‘λ", visible=False) | |
download_button = gr.Button("CSV λ€μ΄λ‘λ", visible=False) | |
gr.Markdown("<h3 style='color: #34495e;'>2. μ§μλ³ μΆμ² νλ‘κ·Έλ¨ λ° μ νλΈ μ½ν μΈ νμΈ</h3>") | |
chatbot = gr.Chatbot() | |
msg = gr.Textbox(label="μ§μ μ΄λ¦μ μ λ ₯νμΈμ") | |
clear = gr.Button("λν λ΄μ μ§μ°κΈ°") | |
# λΆμ λ²νΌ ν΄λ¦ μ ν μ΄λΈ, νμΌ λ€μ΄λ‘λλ₯Ό μ λ°μ΄νΈ | |
analyze_button.click( | |
hybrid_rag, | |
inputs=[employee_file, program_file, youtube_file, title_col, description_col, url_col, upload_date_col], | |
outputs=[output_table, csv_download, download_button] | |
) | |
# CSV λ€μ΄λ‘λ λ²νΌ | |
download_button.click(download_csv, inputs=[], outputs=[csv_download]) | |
# μ±ν κΈ°λ₯ | |
msg.submit(chat_response, [msg, chatbot], [chatbot]) | |
clear.click(lambda: None, None, chatbot, queue=False) | |
# νλ‘κ·Έλ¨ μ’ λ£ μ μμ νμΌ μμ | |
import atexit | |
def cleanup(): | |
global global_csv_file | |
if global_csv_file and os.path.exists(global_csv_file): | |
os.remove(global_csv_file) | |
# Gradio μΈν°νμ΄μ€ μ€ν | |
demo.launch() |