Friend　recommendation　algorithm　plays　an　important　role　in　social　networks.　However,　traditional　recommendation　algorithms　do　not　take　into　account　the　drifting　of　user　interests,　and　there　are　also　some　deficiencies　when　the　recommendation＇s　timeliness　is　considered.　Aimed　at　this　problem,　the　measure　method　of　similarity　was　improved　by　combining　with　the　characteristics　of　user　interest＇s　change　with　time.　A　time　decay　model　was　introduced　to　measure　the　predictive　value.　At　the　same　time,　the　method　refines　the　preference　of　the　target　users　according　to　the　homogeneity　theory　through　the　interest　preference　of　friends.　Through　experiments　on　Sina　microblog　data　set,　and　the　results　show　that　the　algorithm　to　achieve　better　recommendation　effect　on　precision,　recall　and　F　value.　©　2019　IEEE.