Situation　element　extraction　is　a　significant　step　of　network　security　situation　awareness.　This　paper　presents　a　network　security　situation　element　extraction　model　based　on　projection　pursuit　regression.　A　nonlinear　optimization　problem　is　proposed　in　the　model,　in　which　the　projection　direction　and　the　polynomial　coefficients　are　optimization　variables,　and　projective　index　function　is　target　function.　Due　to　the　high　precision　and　fast　convergence　speed,　the　particle　swarm　optimization　is　introduced　to　solve　the　optimization　problem.　The　simulation　experimental　results　show　that　this　model　is　efficient.　©　2012　IEEE.