Would You Find It, Or Shall We Recommend It To You? 🧐✨

4 min readDec 4, 2023

Welcome, since you’re here, I’ll be the one recommending to you 😊

In this article, we’ll take a close look at the techniques used in recommendation systems. Let’s get started!

Recommendation Systems are algorithms or tools that suggest items to users based on their preferences, behavior, or similarities to other users.

These systems analyze available data, such as past interactions, purchases, or ratings, to generate personalized recommendations for users.

They are used in various applications, including e-commerce, streaming platforms, social media, and more, to enhance user experience by offering tailored suggestions or content.

Association Rule Learning

The primary goal of Association Rule Learning is to uncover frequent patterns, co-occurrences, or correlations among items in datasets. This technique is commonly employed in various fields such as market basket analysis, recommendation systems, web mining, and more.

Association Rule Learning algorithms aim to extract rules that describe the relationships between different items in the data. For example, in a retail transaction dataset, an association rule might indicate that if customers buy product A, they are also likely to buy product B.

The Apriori algorithm is one of the popular algorithms used in Association Rule Learning. It identifies frequent item sets by employing measures like support, confidence, and lift to generate rules that reveal the relationships between items.

How does the apriori algorithm work?

We examine items that frequently appear together and calculate their support values. Then, based on a predefined threshold value that we determined earlier, we filter out item combinations with support values below this threshold.

In this process, item sets are repeatedly generated, elimination is performed based on the support value, and the process is repeated until frequent itemsets are found.

Finally, the itemsets are combined, and rule metrics are computed to determine the strength and significance of the association between different items or itemsets. This involves calculating metrics like confidence, support, and lift for the generated association rules.

Content-Based Systems

These systems suggest items to users based on the characteristics or attributes of the items themselves and the user’s historical preferences. For instance, in a content-based movie recommendation system, if a user has previously liked action movies, the system might recommend other action movies based on similar attributes like genre, actors, or directors.

Collaborative Filtering Systems

These systems make recommendations by identifying patterns or similarities between users’ behavior or preferences. They recommend items based on the preferences of similar users. There are two types of collaborative filtering:

User-Based Filtering: This method suggests items that similar users have liked or purchased. For example, if User A has similar preferences to User B and User B has liked a particular movie, the system might suggest that movie to User A.
Item-Based Filtering: This method makes recommendations by leveraging information derived from the general behavior of the community. For example, it suggests other films that resemble the liking structure of a particular film to users.

Model Based Matrix Factorization

Model-based matrix factorization is a technique used in recommender systems and data analysis to decompose a matrix into lower-dimensional matrices that capture underlying patterns or latent factors. The goal is to approximate the original matrix by multiplying these lower-dimensional matrices. The matrix being factorized typically represents interactions or relationships between two types of entities. For instance, in a recommender system, the matrix might represent users and items, with entries indicating user-item interactions (like ratings or preferences).

Let’s look at an example here! Our goal is to fill in the BLANK cell based on latent factors.