With　the　rapid　development　of　cyberspace,　cyber　security　incidents　are　increasing,　and　the　means　and　types　of　network　attacks　are　becoming　more　and　more　complex　and　refined,　which　brings　greater　challenges　to　security　risk　control.　First,　the　knowledge　graph　technology　is　used　to　construct　a　cyber　security　knowledge　graph　based　on　ontology　to　realize　multi-source　heterogeneous　security　big　data　fusion　calculation,　and　accurately　express　the　complex　correlation　between　different　security　entities.　Furthermore,　for　cyber　security　risk　control,　a　key　node　assessment　method　for　security　risk　diffusion　is　proposed.　From　the　perspectives　of　node　communication　correlation　and　topological　level,　the　calculation　method　of　node　communication　importance　based　on　improved　PageRank　Algorithm　and　based　on　the　improved　K-shell　Algorithm　calculates　the　importance　of　node　topology　are　studied,　and　then　organically　combine　the　two　calculation　methods　to　calculate　the　importance　of　different　nodes　in　security　risk　defense.　Experiments　show　that　this　method　can　evaluate　the　importance　of　nodes　more　accurately　than　the　PageRank　algorithm　and　the　K-shell　algorithm.　©　2021　IEEE.