With　the　development　of　network　technology　and　network　services,　the　number　of　Web　services　will　also　be　exploded,　so　the　magnitude　of　Web　services　that　can　provide　similar　functions　will　gradually　be　increased.　Quality　of　service　(QoS)　is　widely　used　to　describe　and　evaluate　the　non　functional　properties　of　Web　services,　and　has　been　successfully　applied　to　service　recommendation.　However,　most　of　the　current　Web　Service　recommendations　are　still　limited　to　traditional　collaborative　filtering,　without　considering　the　bias　information　of　users　and　web　services　themselves.　In　this　thesis,　firstly,　the　author　introduces　the　current　situation　of　web　services　recommendation,　then　analyzes　the　existing　problems　and　the　characteristics　of　the　data　sets.　Finally,　the　author　applies　the　BiasSVD　to　predict　QoS,　and　then　recommends　to　users.　By　comparing　the　experimental　results　with　the　traditional　collaborative　filtering　algorithm,　the　feasibility　and　superiority　of　the　algorithm　in　Web　services　recommendation　scenarios　are　obtained.　©　2020　IEEE.