基于机器学习的求助类警情风险分析

打开文本图片集

中图分类号：D917.6 文献标识码：A 文章编号：2095-6916（2025）15-0060-04

Risk Analysis of Help-Seeking Police Alerts Based on Machine Learning

Du Hao

（Zhejiang Police College，Hangzhou 310053）

Abstract：Trough heriskanalysisofhelp-seekingpolicealerts，riskpredictionandfeatureanalysiscanbeachieved，providingdecision-makingsupportforpublicsecurityorgans.Acomparativestudyisconductedontwodiferentappoachesforriskanalysisofhelpsekingpolicealerts：textbasednds-based.Intetetasedanalsismetod，Word2vcisusdtoextracttextalaturesofpoe alerts，andarndomforestparticlesaoptizatiomodelisusedfotrainngIntheisk-basedanalysisetodarissest indexsystemforelpseekingpolicealertsisonstructedfromfourdimensios：subjectriskhavioralisk，venueriskandeyfactor risk.BERT-MCisusedtoextractriskelatedfeaturesofpolicealerts，andaandomforestpartcleswaroptizationmodelisld fortraining.Experimentalresultsshowthattherandomforest-particleswarmoptimizationmodelbasedonriskfeaturescanbasicalachieve risk prediction and feature weight analysis，demonstrating practical significance.

Keywords：policealerts；help-seeking;risk analysis；artificial intelligence

面对日趋复杂的社会形势，公安机关需要接受处理的警情愈发繁杂，基层往往出现警力资源供求匹配程度不足、供不应求的现象。（剩余5828字）