简介：Anintelligentsecuritysystemsengineeringapproachisusedtoanalyzefireandexplosivecriticalincidents,agrowingconcerninurbancommunities.Afeed-forwardback-propagationneuralnetworkmodelsthedamagesarisingfromthesecriticalincidents.Theoverallgoalistopromotefiresafetyandsustainablesecurity.Theintelligentsecuritysystemsengineeringpredictionmodelusesafullyconnectedmultilayerneuralnetworkandconsidersanumberoffactorsrelatedtothefireorexplosiveincidentincludingthetypeofpropertyaffected,thetimeofday,andtheignitionsource.ThenetworkwastrainedonalargenumberofcriticalincidentrecordsreportedinToronto,Canadabetween2000and2006.Ourintelligentsecuritysystemsengineeringapproachcanhelpemergencyrespondersbyimprovingcriticalincidentanalysis,sustainablesecurity,andfireriskmanagement.

简介：ThispaperfocussesontheinformationasymmetryincrisisnewsafteraseriousincidentinTianjin,China,in2015.Theincidentcausedenormousdamageandresultedinsocietalunrestbecauseofthelackofreliableinformationfromtheformalmediachannels.Socialmedia—microblogs—playedamajorroleinreportingoncrisissituations.Wedividednetizens(i.e.,thecitizensofthenet)intohighandlowtypesaccordingtotheirinformation-criticalleveltothecrisisnews.Thedatashowsinformationdeteriorationonthecrisisnews,relatedtothenetizens’information-criticallevel.Forthetraditionalmediathereistheopportunitytouseinformationqualitydistortiontomakemoremarginalprofits.Thisispossibleonlyifthecitizens’informationstaysunderacertainqualitylevel.Theresultisoverprovisionoflowqualitynewsandhighqualitynewsdrivenoutofthemarket,whereuponadverseselection(i.e.,alackofsymmetricinformation)appears.However,byadoptingaprocessview,wefoundself-correctingmechanism(i.e.,dyingoutofrumors)ofthesocialmediacommunitiesinChina.Weprovidedaagent-basemodelandsimulationtoshowthatthemoremediaexistinthemarket,thefasterspeedoftheinformationdeterioration,butalsothecapacityto‘discuss’rumors.