Sub-projects aimed at conceptualizing, designing, and building recommendation engines, for in whichever sector or domain the need for filtering arises. Also contains traditional logic based algorithmic approaches to filtering. Hybrid models are also present that augment such models with machine learning (ML) and deep learning (DL) approaches.
Hybrid models perform the best, because collaborative filtering suffers from the cold start problem (new users are tough to figure out. Hell! in real life people we know since childhood sometimes act strange 🙃) and content-based filtering suffers from the complex and tough-to-figure-out task of collecting/extracting most relevant features for representing the user's psychy and behaviour and interest regarding that specific product (Afterwards, it's simple, as then we can apply simple clustering/grouping (k-Means) or classification (SVM/Decision-Trees, with classes being the different groups in which multiple user instances in the training data reside) models). So a hybrid model is a sweet-spot between these two.
-
Personalization : Can guide
personalized feedsection. that ultimate drive the growth of the business. -
Targeted Advertisement / Push-Notifications / Deals: Increase revenue and improve busineed via such measures that increase the sale of the perfectly 'synced' items and users.