kmrmanish/LPI_Course_Recommendation_System

1import streamlit as st2import difflib3import pandas as pd4import numpy as np5import re6import nltk7from nltk.corpus import stopwords8from nltk.stem.porter import PorterStemmer9from sklearn.feature_extraction.text import TfidfVectorizer10from sklearn.metrics.pairwise import cosine_similarity11 12# Download NLTK stopwords if not already done13nltk.download('stopwords')14 15# Read the data16lpi_df = pd.read_csv('Learning Pathway Index.csv')17 18# Rename columns19lpi_df.rename(columns={20    "Course / Learning material": "Course_Learning_Material",21    "Course Level": "Course_Level",22    "Type (Free or Paid)": "Type",23    "Module / Sub-module \nDifficulty level": "Difficulty_Level",24    "Keywords / Tags / Skills / Interests / Categories": "Keywords"25}, inplace=True)26 27# Combine features28lpi_df['combined_features'] = lpi_df['Course_Learning_Material'] + ' ' + lpi_df['Source'] + ' ' + lpi_df['Course_Level'] + ' ' + lpi_df['Type'] + ' ' + lpi_df['Module'] + ' ' + lpi_df['Difficulty_Level'] + ' ' + lpi_df['Keywords']29 30# Text preprocessing31combined_features = lpi_df['combined_features']32porter_stemmer = PorterStemmer()33 34def stemming(content):35    stemmed_content = re.sub('[^a-zA-Z]', ' ', content)36    stemmed_content = stemmed_content.lower()37    stemmed_content = stemmed_content.split()38    stemmed_content = [porter_stemmer.stem(word) for word in stemmed_content if not word in stopwords.words('english')]39    stemmed_content = ' '.join(stemmed_content)40    return stemmed_content41 42combined_features = combined_features.apply(stemming)43 44# TF-IDF and similarity45vectorizer = TfidfVectorizer()46vectorizer.fit(combined_features)47combined_features = vectorizer.transform(combined_features)48similarity = cosine_similarity(combined_features)49 50# Streamlit app51st.title('Learning Pathway Index Course Recommendation')52user_input = st.text_input('Enter What You Want to Learn : ')53 54if user_input:55    list_of_all_titles = lpi_df['Module'].tolist()56    find_close_match = difflib.get_close_matches(user_input, list_of_all_titles)57 58    if find_close_match:59        close_match = find_close_match[0]60        index_of_the_course = lpi_df[lpi_df.Module == close_match].index.values[0]61        similarity_score = list(enumerate(similarity[index_of_the_course]))62        sorted_similar_course = sorted(similarity_score, key=lambda x: x[1], reverse=True)63 64        st.subheader('Courses suggested for you:')65        for i, course in enumerate(sorted_similar_course[:30], start=1):66            index = course[0]67            title_from_index = lpi_df.loc[index, 'Module']68            st.write(f"{i}. {title_from_index}")69 70        if len(sorted_similar_course) == 0:71            st.write('No close matches found.')72    else:73        st.write('No close matches found.')74