app.py

import pandas as pd
from sentence_transformers import SentenceTransformer, util
from transformers import pipeline
import torch
import gradio as gr
import os

# Load the dataset
csv_file_path = os.path.join(os.getcwd(), 'Analytics_Vidhya_Free_Course_data.csv')
df = pd.read_csv(csv_file_path, encoding='Windows-1252')
df.fillna('', inplace=True)

# Load the pre-trained model for embeddings
model = SentenceTransformer('multi-qa-mpnet-base-dot-v1')

# Combine title and description to create a full text for each course
df['full_text'] = df.iloc[:, 0] + " " + df.iloc[:, 1] + " " + df['Instructor Name'] + " " + df['Rating'].astype(str) + " " + df['Category']

# Convert full course texts into embeddings
course_embeddings = model.encode(df['full_text'].tolist(), convert_to_tensor=True)

# Load a model for text generation (e.g., BART)
generator = pipeline('text2text-generation', model='facebook/bart-large-cnn')

def expand_query(query):
    paraphraser = pipeline('text2text-generation', model='Vamsi/T5_Paraphrase_Paws')
    expanded_queries = paraphraser(query, num_return_sequences=3, max_length=50, do_sample=True)
    return [q['generated_text'] for q in expanded_queries]

def generate_description(query):
    response = generator(query, max_length=100, num_return_sequences=1)
    return response[0]['generated_text']

def search_courses(query, level_filter=None, category_filter=None, top_k=3):
    expanded_queries = expand_query(query)
    all_similarities = []

    for expanded_query in expanded_queries:
        query_embedding = model.encode(expanded_query, convert_to_tensor=True)
        similarities = util.pytorch_cos_sim(query_embedding, course_embeddings)[0]
        all_similarities.append(similarities)

    aggregated_similarities = torch.max(torch.stack(all_similarities), dim=0)[0]
    filtered_df = df.copy()

    if level_filter and level_filter != "Nil":
        filtered_df = filtered_df[filtered_df['Level of Difficulty'] == level_filter]
    if category_filter and category_filter != "NIL":
        filtered_df = filtered_df[filtered_df['Category'] == category_filter]

    if filtered_df.empty:
        return "<p>No matching courses found.</p>"

    filtered_similarities = aggregated_similarities[filtered_df.index]
    top_results = filtered_similarities.topk(k=min(top_k, len(filtered_similarities)))

    results = []
    for idx in top_results.indices:
        idx = int(idx)
        course_title = filtered_df.iloc[idx]['Course Title']
        course_description = filtered_df.iloc[idx, 1]
        course_url = filtered_df.iloc[idx, -1]
        generated_description = generate_description(course_title + " " + course_description)
        course_link = f'<a href="{course_url}" target="_blank">{course_title}</a>'
        results.append(f"<strong>{course_link}</strong><br>{course_description}<br>{generated_description}<br><br>")

    return "<ol>" + "".join([f"<li>{result}</li>" for result in results]) + "</ol>"

def create_gradio_interface():
    with gr.Blocks() as demo:
        gr.Markdown("#  Analytics Vidhya Free Courses")
        gr.Markdown("Enter your query and use filters to narrow down the search.")
        query = gr.Textbox(label=" Search for a course", placeholder="Enter course topic or description")
        with gr.Accordion(" Filters", open=False):
            level_filter = gr.Dropdown(choices=["Beginner", "Intermediate", "Advanced", "Nil"], label=" Course Level", multiselect=False)
            category_filter = gr.Dropdown(choices=["Data Science", "Machine Learning", "Deep Learning", "AI", "NLP", "NIL"], label=" Category", multiselect=False)
        search_button = gr.Button("Search")
        output = gr.HTML(label="Search Results")
        search_button.click(fn=search_courses, inputs=[query, level_filter, category_filter], outputs=output)

    return demo

# Launch Gradio interface
demo = create_gradio_interface()
demo.launch()