基于协同过滤和社交网络挖掘的个性化微博推荐

《基于协同过滤和社交网络挖掘的个性化微博推荐》由会员上传分享，免费在线阅读，更多相关内容在教育资源-天天文库。

1、ᜧᦻឋ!"#$：ᡠᕈḄᦻ，*ᙠ,-Ḅᢣ,/，0123Ẇ567ᡠ89Ḅᡂ;。◀ᦻ>?@A$BᵨḄᑁE᜜，ᦻGᒹIJKᐸMNᡈPQR@S⊤ᡈUᑏWḄ7Xᡂ;。YᦻḄẆZ["⌕^_ḄN`PQ，ᙳRᙠᦻ>b$cde᪗$。gᐰijᑮ#$Ḅlmn;ᵫpq。ᦻ7ὅstuv：wIᨴIu万方数据ᜧᦻ᱐ᩗᵨᣴᩗᦻ7ὅgᐰyz᪥ᨵᐵ~ᶇ、ᵨᦻḄ，i᪥~ᶇᔣᨵᐵᡈ᪀〈ᦻḄᓺ`ᵯ᱐，ᐕᦻ▅`▅。ᣴᩗᜧbᦻḄᐰᡈᑖᑁEᐭᨵ

2、ᐵᦪ ¡23¢£，b¤ᵨ¥ᓺ、¦ᓺᡈ§¨©ᑴ«¬~`®ᦻ。~¯□，ᙠ__wz¯°〉ᵨᣴᩗ²。ᦻ³´;G~¯g/(ᙠbd᪾ᑁᡭ“v”）ᦻ7ὅst:ᢣ,ᦟ-st¹vwI万方数据᢬᣾Ḅ!ឋᓄ#$%₞᤺⌕º»¼，½¾¿º»，ÀÁº»，ᙠÂw?@ÃÄᡂÅÆÇÈÉ3ḄÊËÌÍÎ，ÏᜩÆÑὶËᵨᡝÔÕᜧÖḄ×ØÙÚ。º»ÛÚbØÜÝÞ23᣸ᑡ，á᪵ᵨᡝãbᦈᑮMᡠᐵAḄSåḄ×ØÛÚ，æ>çèRéᐶëḄ23▅ì。í*，ÙÚWîḄï⚪ñº»ᵨᡝòᩭyôõö᡾，øᐸ*

3、ùúᨵôõᐵAḄᵨᡝ，ÏᜩûᦈᑮᡂᓟýḄº»ÛÚ。ᦻᐵA´᝞Kᔣᵨᡝÿ₞MéᐶëḄº»，æMçᑮèRéᐶëḄÛÚᡠ◤⌕Ḅ@ᔊ。º»ḄôõÛÚbᵨ´7Æᔣᵨᡝÿ₞ᱯº»ÛÚḄ ，ᒹᵨᡝḄSå，S，Êᐵ©。ᦻÔyN´W`ÊËÌ᢬᣾Ḅdlᩭ9ᵨᡝḄNᐶëæ23ÿ₞。᜜，áNdlbᨵᦔᙢnᔠᐸMᨵᵨḄ/ᦻÙÚ。ᦻᨬ$Ḅdl%⌕ὃ⇋yº»☢Ḅ)d☢*+：º»ÛÚḄ,⚪*+，ᵨᡝÊᐵḄ*+，b-ᐸM.eᱯ/，½º»ÛÚSåὅḄᩗ0

4、12，º»ÙÚ3Ḅ4Ö©。×5.6，áú*+Y´º»ÙÚḄÿ₞ÇÈ"⌕，7ᦻdlḄᦔ;8WyᐸMõ9l。ᐵ,-：º»ÿ₞WNឋᓄÊËÌ——i——万方数据万方数据PersonalizedMicroblogRecommendationbasedonCollaborativeFilteringandSocialNetworkMiningABSTRACTMicroblogs,suchSinaWeibo,TencentWeibohaverapidlygrownintopopularsoci

5、alnetworkservicesrecently.Theyprovideslargenumbersofreal-timemessagesforusersandusercangetfreshinformationbyscanningtheirtimeline.Microblogmessagesarepresentedinchronologicalorderandusersscanthefollowees’timelinestofindwhattheyareinterestedin.However,

6、aninformationoverloadproblemhastroubledmanyusers,especiallythosewithmanyfolloweesandthousandsoftweetsarrivingeveryday.Inthispaper,wefocusonrecommendingusefultweetsthatusersarereallyinterestedinpersonallytoreducetheusers’efforttofindusefulinformation.M

7、anykindsofinformationonTwitterareavailableforhelpingrecommendation,includingtheuser’sowntweethistory,retweethistoryandsocialrelationsbetweenusers.Weproposeamethodofmakingtweetrecommendationsbasedoncollaborativefilteringandsocialnetworkminingtocapturep

8、ersonalinterests.Itcanalsoconvenientlyintegratetheotherusefulcontextualinformation.OurfinalmethodconsidersthreemajorelementsonMicroblogs:tweettopiclevelfactors,usersocialrelationfactorsandexplicitfeaturessuchasauthorityofthepublisherandquality