Optimization and application of intelligent recommendation algorithm in emergency public welfare advertisement personalized pushing system
LIU Dongmei
ZHONG Jianyu
ZHANG Hongnuo
WANG Yang
Abstract：In emergency scenarios,the timely and accurate push of emergency public service advertisements to the public is of great significance for enhancing public emergency awareness,guiding correct emergency behav-iors,and ensuring social security and stability.Traditional intelligent recommendation algorithms face numerous challenges when applied to the push of emergency public service advertisements.For example,these challenges include data sparsity and cold start dilemma in collaborative filtering algorithms,optimization of ran-dom forest and support vector machine algorithms in complex emergency scenarios,etc.This paper proposed an emergency-scenario-driven dynamic fusion and adaptive recommendation model of multiple algorithms.It in-novatively integrates collaborative filtering,random forest,and support vector machine algorithms,and incorpo-rates strategies such as data preprocessing,the introduction of emergency scenario context information,and on-line learning.The model effectively solved the problems of traditional algorithms in the push of emergency public service advertisements,and theoretically provided a new algorithm framework for the personalized push system of emergency public service advertisements.Verified by numerous experiments and practical cases,the optimized algorithm has achieved significant improvements in key indicators such as recommendation accuracy,user engagement,and advertisement conversion rate.The technology provides strong technical support for the development of personalized emergency public service advertising push systems.
Keywords：intelligent recommendation algorithmpersonalized pushing of emergency public welfare advertise-mentscollaborative filteringrandom forestsupport vector machine
Publication Date：2025-08-30
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：11( 98-108 )