File size: 1,778 Bytes
562afe3 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 |
import pandas as pd
import numpy as np
import nltk
from nltk.stem.snowball import SnowballStemmer
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.metrics.pairwise import cosine_similarity
import streamlit as st
from PIL import Image
# Load the dataset
data = pd.read_csv('amazon_product.csv')
# Remove unnecessary columns
data = data.drop('id', axis=1)
# tokenizer and stemmer
stemmer = SnowballStemmer('english')
def tokenize_and_stem(text):
tokens = nltk.word_tokenize(text.lower())
stems = [stemmer.stem(t) for t in tokens]
return stems
# stemmed tokens column
data['stemmed_tokens'] = data.apply(lambda row: tokenize_and_stem(row['Title'] + ' ' + row['Description']), axis=1)
# TF-IDF vectorizer and cosine similarity function
tfidf_vectorizer = TfidfVectorizer(tokenizer=tokenize_and_stem)
def cosine_sim(text1, text2):
# tfidf_matrix = tfidf_vectorizer.fit_transform([text1, text2])
text1_concatenated = ' '.join(text1)
text2_concatenated = ' '.join(text2)
tfidf_matrix = tfidf_vectorizer.fit_transform([text1_concatenated, text2_concatenated])
return cosine_similarity(tfidf_matrix)[0][1]
# search function
def search_products(query):
query_stemmed = tokenize_and_stem(query)
data['similarity'] = data['stemmed_tokens'].apply(lambda x: cosine_sim(query_stemmed, x))
results = data.sort_values(by=['similarity'], ascending=False).head(10)[['Title', 'Description', 'Category']]
return results
# web app
img = Image.open('download.png')
st.image(img,width=600)
st.title("Intelligent Product Finder for Amazon")
query = st.text_input("Enter Product Name")
sumbit = st.button('Search')
if sumbit:
res = search_products(query)
st.write(res) |