题目：基于情感感知的个性化音乐推荐技术研究英文论文题目：Researchonpersonalizedmusicrecommendationtechnologybasedonemotionperception摘要在音乐举荐方面，目前大部分的举荐体系仍然以固有的办法为基础，情境感知在特色化的音乐举荐方面的使用仍然处在开始环节。固有的举荐办法通常仅仅重视“客户-项目”间存在的关联关系，很少思考客户行为产生的时候对照的情境数据（比如时间、天气等相关部分）。以情境感知为基础的特色化音乐举荐，举荐经过把情境数据引进举荐计算方法，不但可以符合客户特色化的举荐需要，能够从大量音乐库内协助客户迅速定位至符合本身喜爱的音乐，另外可以较佳地展现普适核算的价值，也就是数据与核算资源的获得和连入能够出现在任意时间、任何地点、通过任意模式。这篇文章探究的情境感知举荐方法，关于目前音乐举荐方面具备某一程度的真实参照价值这篇文章从选题探究的有关环境着手，先是陈述音频划类和检查搜索技术的前提，之后具体的剖析纯音乐和语音-音乐混合型片段的高准确识别算法，接着又探析了基于MFCC(Mel-FrequencyCepstralCoefficients)和GMM(Adaptivebackgroundmixturemodelsforreal-timetracking)的歌曲个性计算方法，最后重点的研究了歌曲个性计算与推荐原型系统设计与实现。关键词：音频检索；歌曲；个性计算；音乐推荐AbstractInthefieldofmusicrecommendation,mostofthecurrentrecommendationsystemsarebasedontraditionalmethods,andtheapplicationofsituationalawarenessinpersonalizedmusicrecommendationisstillinitsinfancy.Thetraditionalmethodsofrecommendationtendtofocusonlyonthetwodimensionalrelationshipbetweenthe“userproject”andlessconsiderationofthesituationalinformation（suchastime,place,weather,etc.）whentheuser'sbehavioroccurs.Thepersonalizedmusicrecommendationbasedonsituationalawarenessisrecommendedbyintroducingthesceneinformationintotherecommendationalgorithm.Ontheonehand,itcannotonlymeetthepersonalizedrecommendationrequirementsoftheusers,thatis,itcanhelptheuserstoquicklylocatethemusictosatisfytheirownpreferencesfromthemassmusiclibrary,ontheotherhand,itcanalsoreflectthevalueofpervasivecomputing,thatis,thevalueofthepervasivecomputingismoreimportant.Accessandaccesstoinformationandcomputingresourcescanoccuratanytime,anywhere,inanyform.Thescenarioawarerecommendationalgorithmproposedinthispaperhassomepracticalreferencevalueforthecurrentmusicrecommendationfield.Startingwiththerelevantbackgroundoftheresearch,thispaperfirstexpoundsthebasisofaudioclassificationandretrievaltechnology,thenanalyzesthehighaccuraterecognitionalgorithmofpuremusicandspeechmusicmixedfragments,andthenanalysesthemethodofsongpersonalitybasedonMFCC(Mel-FrequencyCepstralCoefficients)andGMM(Adaptivebackgroundmixturemodelsforreal-timetracking),andfinallystudiesthecharactercalculationandrecommendationofthesong.Thedesignandimplementationofthetypesystem.Keywords:AudioRetrieval;Song;PersonalComputing;MusicRecommendation目录摘要........................................................................................................................................................IAbstract...................................................................................................................................................II目录.....................................................................................................................................................III第1章绪论...............................................................................................................................................11.1研究背景........................
VIP
