泰迪杯數(shù)模優(yōu)秀論文-網(wǎng)絡(luò)招聘信息的數(shù)據(jù)挖掘與綜合分析

1、第四屆“泰迪杯”全國(guó)數(shù)據(jù)挖掘挑戰(zhàn)賽優(yōu)秀作品作品名稱(chēng)：網(wǎng)絡(luò)招聘信息的數(shù)據(jù)挖掘與綜合分析榮獲獎(jiǎng)項(xiàng)：特等并獲企業(yè)冠名獎(jiǎng)作品單位：北京林業(yè)大學(xué)作品成員：孫海鋒鄭中樞楊武岳指導(dǎo)老師：崔曉暉第四屆“泰迪杯”全國(guó)數(shù)據(jù)挖掘挑戰(zhàn)賽www.tipdm.g第2頁(yè)wkRecruitmentInfmationAnalysisMiningAbstractInrecentyearswiththewideapplicationofIntertherapiddevelo

2、pmentofInterrecruitmentrecruitmentinfmationwkplatfmhasbecomethemainchannelfinterviewerstoobtaininfmation.Therefeusingthewktextanalysisdataminingtechnologytowkrecruitmentinfmationoftheresearchisofgreatsignificance.Aiminga

3、ttheproblemofthefirsttherecruitmentinfmationtablebyPositionIdjobdeiontabletoheavydontrepeatjobinfmation.UsingjiebaChinesewdsegmentationtoolstoparticipleofjobdeioninfmationthroughtheTFIDFalgithmtoextracteachjobdeionofthet

4、opfivekeywds.ReusingtheTFIDFalgithmfeachjobdeionoftheTFIDFweightvecttheKmeansofTFIDFweightvectclusteringgetsevencentersofmass.Respectivelycalculatedfromthecenterofmassofrecent5positioncombinationofrecruitmentinfmationtab

5、lePositionFirstTypefieldsbasedonKNNalgithmfeachclasswithnatureoftheindustry.ThenrespectivelythestatisticalanalysisofvariousprofessionaltypesofPositionNamedrawingthecrespondingprofessionalcareertype.Aimingattheproblemofth

6、esecondbyusingexceltogoafterheavyrecruitmentinfmationtable(IndustryField)toindustryfieldwkarea(City)thepositionclassification(PositionFirstType)classifythreeprojectsionallkindsofcontenttocalculatesummarystatisticsfeachpr

7、ojectaccdingtocalculatingmethancontenttoindustryregionduetobepopularposition.Aimingattheproblemofthethirdaccdingtotheactersofthepositionofdatamininganalysisdefinsingthepositionofemergingastwocategiesfilteredrespectively.

8、UsingdivergentthinkingthenwouldgetresultsofscreeningoutrespectivelyaccdingtotheCity(City)phase(Financestage)Education(Education)compensation(Salary)fromfouraspectsonthevariousstatisticssystematicallywiththedemoftanalysis

9、fecastrelatedposition.Aimingattheproblemofthefthbylookingftheitpositioncrespondingtotheidofthejobdeionthewdsegmentationbuiltitprofessionalsemanticlibraryonthebasisofscreeningallitposition.Toannex1fdatapreprocessingindata

10、preprocessingthedataonapreliminaryscreeningthepositionoftheitindustry.Toscreenouttheitpositionsthecrespondingprofessionalidtofindthejobdeiontableinthejobdeiondeiononthebuilditprofessionalsemanticreposity.Determiningthejo

11、bdeioninthetablepositionisinlinewiththeitprofessionthroughthejudgmentprofessionalsemanticrepositytodeterminewhethertheintersectionoflengthftheitprofessionalstatisticalarea.Aimingattheproblemofthefifthaccdingtotheresearch