Create Restful API by using ML/DL to find/save/update same products in different e-commerce websites.

I need to have a RESTful API to scraping some websites and find prices for same product. All websites you need to scrap located in us. We have Amazon, Best Buy, Newegg. Scenario: This API find a product in Amazon (Coca-Cola 1.5 lt sugar free). Now, API should looking for this product in Best Buy and Newegg as well. When same product found in other e-commerce websites (By using ML explained below section), API should save the product name and all prices in MongoDB collection. Name, prices, and images addresses of the product in all scraped e-commerce websites should save in MongoDB. All image addresses of product in all website should store in MongoDB collection separated by website name. For example related to above line: We have 3 e-commerce websites. Scraped product has 5 images in every online shopping websites. In this case we need to save 15 image address in MongoDB. Every photo should have an attribute (in MongoDB) to define the online shopping website that image belongs to. This web app should scrap the whole website and do same job for all products. This task should run every hour. App should use Machine Learning to make sure of same product finding in different websites Need to introduce the Deep Learning models to figure out the same products based on product pictures or product names and descriptions. App should compare some part of product listing in different e-commerce websites. * Product Title * Product Description * Product Images We should be able to define similarity score for ML to consider that score and above as same product. Example: We will specify 80% similarity for title, 60% for description, 60% for images. If ML meet above scores, It will consider the found product as the same. ML should be improvable. Means, Based on below explanation, We should have a function in API that user can tell us that whether result is correct or not. If user stated YES, we can continue with above score value. if user stated NO, ML should increase the score by 5 (80+5, 60+5, 60+5) WE PREFER TENSORFLOW Main website that you need to start finding a product from, is Amazon. We can define the product category of amazon during development as hard-coded to avoid crawling whole website. We should make sure of same product price comparison from different websites. As you know, Title, images, and description of product can be different in different websites Most important point is finding same product in different website. If we compare two different product, it will be useless. When app found same product in other e-commerce websites, need to store all price + shipping fee, title and images addresses from all compared e-commerce sites in MongoDB. App should response the product price sorted by price from low to high included shipping fee as API function response as JSON. All images addresses will be response for every website of product by using unique id of website which saved into the MongoDB collection for every product. We should have a function in API to rate the quality of product finding. Means, When user see the product price comparison, can tell us that app did right or not. If not, We need to improve DL concept (explained section above). Products should be categorized by using Ai. App should categorize the products in right category automatically looks like as below example. -Electronic * Laptop and PC ** Desktop PC *** Gaming PC *** Daily Usage PC ** All-in-One PC ** Laptop *TV ** LED ** LCD ** QLED * Air Conditioner FIRST OF ALL, READ PROJECT DETAILS AND TELL ME HOW YOU WANT TO DO THIS JOB ? HOW YOU WANT TO IMPLEMENT ML/DL ? WHAT IS YOUR PLAN TO AVOID BLOCKING THE APP FROM CRAWLING BY E-COMMERCE WEBSITES WHILE WE NEED TO UPDATE ALL PRODUCT PRICES MANY TIMES IN A DAY ? MEANS, WE NEED TO FIND NEW PRODUCT AND UPDATE CURRENT PRODUCT DETAILS IN DB MANY TIME IN A DAY