The　personalized　recommending　system　approaches　have　been　widely　been　employed　in　e-commerce　to　help　users　find　items　they　like.　Recommender　algorithm　is　the　core　of　the　personalized　recommending　system.　The　Slope　One　algorithm　is　a　commonly　used　collaborative　filtering　algorithm.　It　is　a　much　simpler　algorithm　that　not　only　easy　to　maintain　but　also　easy　to　extende.　But　it　do　not　adequately　consider　the　user　similarity　and　item　similarity.　This　paper　proposes　a　new　algorithm　to　improve　its　drawbacks.　The　algorithm　add　the　user　similarity　and　item　similarity　as　the　weighting　factors.　The　Experimental　analysis　on　MovieLens　datasets　shows　that　the　improved　algorithm　can　get　a　better　prediction　accuracy　and　have　a　better　constringency　speed.