Maximizing Cross Sells By Optimizing Similar Item Recommendation

Recommendation systems in e-commerce have been researched extensively in the past decades and it has advanced in parallel with the general progress of technology, in particular with the growth of the web. Recommendations account for the majority of the site’s earnings because they keep visitors connected with the site content for a longer period. The key obstacle here is anticipating the user's intent based on their online behaviour patterns history. For instance, an entrepreneur can entice a user by offering a discount on a specific element depending on the user's online profile data and behaviour history; or an online travel agency can promote a specific flight destination based on the user's most recent query or information demand. Machine learning is generally used to identify these intentions. Customization systems benefit both businesses and consumers: users receive more pertinent orders and reach their purchasing targeted goals, while businesses can boost revenue and average basket capacity. Using strategies spanning from collaborative filtering to learning sequence by producing embedding for products, this research emphasises various issues in providing excellent product suggestions to users in real-time that are peculiar to the scope and variety of the e-Commerce domain data.