AI-Based Recommendation System For Indonesia's E-Commerce

YOGYAKARTA The recommendation system applied to e-commerce has become a common thing. In fact, the system is quite important because it indirectly helps consumers solve problems over their confusion in choosing products. However, the recommendation system must continue to be addressed for the sake of more detailed and personalized output with consumers. From there, the AI-based recommendation system for e-commerce in Indonesia is something to consider.

The AI Recommendation System or AI Product Recommendation System is a system that combines Artificial Intelligence (AI) and Machine Learning (ML) to produce predictions of which products are in demand by consumers in the market place. The system works by analyzing customer data in real and massive times.

AI and ML will process data such as purchasing history, clickstream ( mouse movement), browsing duration in a page, geographical location, interaction with content, and much more. The data is then analyzed to produce a pattern. From this paper, behavior and even consumer interest can be learned.

In a study published in the Computer Journal of Computer Information Technology, Computer Systems (JUKTISI), it was stated that the level of accuracy provided by the AI system could be more detailed. The patterns hidden in e-commerce user data will continue to be studied by machine learning and deep learning.

The more accurate the AI system provides recommendation products to customers, the higher the chance that users are loyal to online shopping platforms. Not only that, sales conversion will also increase and the e-commerce position in the competition between market places will also be stronger.

In providing product recommendations to customers, the way AI works is quite complex and complicated. However, there are at least three concepts applied by AI in recommending the product, namely as follows.

This system is done by studying what products are liked or purchased by similar people like you. For example, you are the type of person who likes to fish. On the other hand, people who like you buy products A, B, and C. AI will recommend products A, B, C to you.

Content-Based Filtering is a content similarity-based recommendation method or product characteristics that have been consumed by users. AI will analyze the attributes of products that have been purchased, viewed, or liked by users, then recommend other products with similar characteristics.

For example, if you buy a type A phone, the system will recommend other products that have feature similarities or relevance, such as softcases, chargers, or phone accessories that match that type.

Hybrid Approach is a system that implements a combination between Collaborative Filtering and Content-Based Filtering. As a result, product recommendations will be more accurate.

That's information related to the AI-based recommendation system for Indonesian e-commerce. Visit VOI.id to get other interesting information.