运输安全风险源辨识模型［J］．西南交通大学学报，2021，56(1):147－152.LUOWH，CAIFT，WUCN，etal．Text-miningbasedrisksourceidentificationmodelfortransportationsafety［J］．JournalofSouthwestJiaotongUniversity，2021，56(1):147－152.ModeratingeffectanalysisofinformationintheprocessingphaseonexpresswaytrafficaccidentdurationCHENJiao-na1，2，JINYin-li3，TAOWei-jun1，LIDao-feng1(1SchoolofElectronicEngineering，Xi’anShiyouUniversity，Xi’an710065，China;2NationalEngineeringLaboratoryForSurfaceTransportationWeatherImpactsPrevention，BroadvisionEngineeringConsultants，Kunming650200，China;3SchoolofElectronicsandControl，Chang’anUniversity，Xi’an710061，China)Abstract:Sinceemergencyresponseisveryimportanttotheevolutionofthedurationinexpresswaytrafficaccidents，itissignificanttostudytheinfluenceofinformationonthedurationintheprocessingphase．Themoderatingeffectmodelofexpresswayaccidentdurationwasestablishedwithstructureddataandtextdata，giventheindexesoftheprocessingphaseinformationasmoderatingvariables．Takingtheinfluencingfactorsasindependentorcontrolvariables，andthedurationasadependentvariable，themodelwasusedtoanalyzethemechanismofprocessingphaseinformationindurationevolution．Undertheviewofthewordfrequency，reporttimes，andtextcapacity，theevaluationindexesweredefinedusingacontentanalysismethod，astenpotentialmoderatingvariableswereproposed．Themoderatingeffectwastestedbytheinteractionswiththeindependentvariableandthemoderatingvariable，andthetypeofmoderatingeffectwasdiscussedbysimpleslopeanalysis．Therobustnesstestverifiedthevaliditybychangingthecontrolvariablesinthemodel．Undertheconfidencelevelof0.01，itisshownthatsignificantmoderatingvariableshavebeenobtainedontherelationshipbetweenaccidentrangeanddurationrespectively，includingZ1，Z2，Z4，Z5，Z6，Z7，Z8，Z9，andZ10．Inotherwords，theninemoderatingvariableshavesignificantdifferencesintheimpactatdifferentlevels．Afterthemoderatingeffecthasbeencalculated，thevalueofZ8，Z6，andZ7is0.049，0.034，and0.032respectively...
