Optimization of artificial intelligence recommendation algorithm based on knowledge graph and machine learning: Computer thermal optimization

Min Li

doi:10.1016/j.tsep.2025.103390

Return

Optimization of artificial intelligence recommendation algorithm based on knowledge graph and machine learning: Computer thermal optimization

2025-03-01

PRE

黎

黎珉 (Min Li) *

DOI：10.1016/j.tsep.2025.103390

Original

Original request for help

Save

Abstract

En 中文

In computer systems, with the continuous improvement of processor performance, the heat generated also increases, which not only affects the performance and life of the computer, but also may lead to energy waste. Traditional thermal optimization methods rely on fixed rules and preset parameters, which are difficult to adapt to dynamic workload and environmental conditions. The objective of this paper is to develop a recommendation algorithm based on knowledge graph and machine learning to achieve dynamic optimization of computer thermal energy. This paper constructs a knowledge graph with computer thermal management knowledge, which integrates the thermal characteristics of different hardware components, workload patterns, and environmental factors. Then, machine learning algorithm is used to analyze the data in the knowledge graph to explore the potential laws of heat management. On this basis, a recommendation algorithm is developed, which can dynamically adjust the thermal energy management strategy of the computer system according to real-time data and prediction results. The experimental results show that the recommendation algorithm based on knowledge graph and machine learning performs well in computer thermal energy optimization. Compared with traditional methods, the new algorithm can predict heat demand more accurately and provide more effective optimization recommendations. The algorithm shows higher efficiency and better scalability when dealing with large-scale data.

Keywords:

Knowledge graph

Machine learning

Artificial intelligence recommendation algorithm

Computer

Thermal optimization