E-commerce recommender system based on improved K-means commodity information management model

Area of Science:

• Computer Science
• Artificial Intelligence
• E-commerce Technology

Background:

• Rapid advancements in internet and smartphone technology have fueled e-commerce growth.
• Current e-commerce recommendation systems lag behind, impacting efficiency and accuracy.
• There is a need for improved recommendation algorithms to meet evolving e-commerce demands.

Purpose of the Study:

• To develop an enhanced K-means clustering algorithm for e-commerce recommender systems.
• To improve the efficiency and accuracy of commodity information management in e-commerce.
• To provide a more effective recommendation solution for the rapidly growing e-commerce industry.

Main Methods:

• Integration of the K-means clustering algorithm with a genetic algorithm.
• Implementation of genetic algorithm encoding, initial population setup, and fitness function definition.
• Management of commodity information using the enhanced K-means clustering approach.

Main Results:

• The enhanced K-means algorithm achieved a recommendation accuracy of 91.1%.
• This accuracy surpasses traditional K-means (87.9%) and fuzzy C-means (84.8%) algorithms.
• The enhanced K-means algorithm demonstrated a 44% faster convergence rate than traditional K-means.

Conclusions:

• The refined K-means clustering algorithm significantly boosts recommendation proficiency and precision in e-commerce.
• This enhanced algorithm offers a superior alternative to existing recommendation techniques.
• The research contributes to advancing the e-commerce industry through improved recommendation technology.