基于案例推理和规则推理的警情智能化决策模型研究

打开文本图片集

中图分类号：TP31 文献标识码：A 文章编号：1006-8228（2025）08-30-04

Abstract：Thetraditionalapproachtohandlingpolicesituationsmainlyreliesonmanualprocessng，whichhasproblemssuchas loweffciencyandstrongsubjectivityToimprovetheeficiencyofhandlingemergencysituations，thisstudyconstructedan inteligentauxiliarydecision-makingmodel.Using6O7milionpolicesituationdatafromacertainprefecture-levelcityfrom2017 to2023，theBERT-MCmodelisusedtoextractsuicide-relatedpolicesituationatributes，andacaselibraryandulelibraryare established.AdoptingtheTF-IDFweightedimprovedKNNalgorithmforcaseretrieval.Theexperimentshowsthattheimproved algorithmsignificantlyimprovesthesimilarityofcasematchingcomparedtothetraditionalKNN.Thismodelcanprovidedecision support for inexperienced police officers.

KeyWords：InteligentDecisio-making;Case-basedReasoning;Rule-basedReasoning;BERT-MRC;TF-IDFWeightedKNNAlgorit

0引言

求助类警情作为社会治安管理中的重要组成部分，能否准确、高效处理直接关系到人民群众的生命财产安全和社会稳定。（剩余6361字）