欢迎来到天天文库
浏览记录
ID:33762495
大小:4.30 MB
页数:70页
时间:2019-03-01
《个性化营养菜谱推荐方法的分析-(5418)》由会员上传分享,免费在线阅读,更多相关内容在行业资料-天天文库。
1、万方数据ABSTRACTInrecentyears,theInteractisgraduallyoverthrowingthetraditionalindustrieswithbusinessmodesrepresentedby“OnlineToOffiine(020)”.Manytraditionalindustriesstarttoprovideuserswithhigh-qualityandmorepersonalizedservicesbyusingtheInteract.Forexam
2、ple,arestaurantcouldlargelyimproveusers’experienceandincreaseitscompetitiveness,whichsupports020modeandprovidescustomerswithnutrientandpersonalizedmenus.ThesystembuiltbypersonalizednutrientreciperecommendationinthisthesisisexactlyasystemwhichCaneasil
3、yprovideusers、析Ⅱ1convenientpersonalizednutrientrecipe.Theriseofonlinemealorderingmakesthesystemofgreatapplicationvalue.Thepersonalizednutrientreciperecommendationsysteminthisthesisincludesthreeparts:thefrontdesktoshowtheDemo,thecorealgorithmsandbacke
4、nddatabase.Thecorealgorithm,namelypersonalizednutrientreciperecommendationmethod,includesdataacquisitionalgorithm,nutritionarrangealgorithmandpersonalizedrecommendationalgorithm.Thisthesiscardedoutresearchonthedesignofsystemandcorealgorithm,themainwo
5、rkisasfollows:1.Puttingforwardafocusedwebcrawlerbasedondouble-queuesortingandmodelself-learningtocrawlIntemetinformationrelatedtoaspecifictopic.Basedontheexistingfocusedwebcrawlerframework,thecrawlerusesdoubleorderedqueuesanddepthattributeoftheURLtoe
6、nhancetheperformanceofsearchstrategymodule,andusesthebloomfiltertooptimizethemoduleofremovingduplicateURL(uniformResourceLocator).Thedataanalysismoduleisoptimizedbyseparatecalculationofthetopiccorrelationofparent—pageandsubpage.Anewmoduleofmodelself-
7、learningisaddedaswell.ThenwegivesthecomparativeexperimentsofthiscrawlerwiⅡlthegeneralwebcrawlerandthedoublequeuetopicwebcrawler,andtheresultshowsthattheharvestRateofthiscrawlerismuchhigher.2。PuttingforwardageneticalgorithmbasedonLdominationandteamcom
8、petitiontosolvemulti-objectiveoptimizationproblemsinhigh·dimensionalspace.Thisalgorithmputsthenon-dominatedsortinggeneticalgorithrnIIasthebasicframework,andusesLdominatedmethodandgroupcompetitionbasedoncrowdingdistancetogetnon-dominatedsorting,select
此文档下载收益归作者所有