In　order　to　enhance　the　service　quality　of　network　community,　satisfy　the　growing　demand　of　community　users,　this　paper　analyzes　the　content　of　network　community　from　different　aspects　and　provides　a　comprehensive　method　to　divide　the　network　community　user　group.　On　one　hand,　this　method　use　link　analysis　techniques　to　study　hyperlink,　calculate　the　number　of　output　links　and　input　links,　and　then　construct　the　diagram　of　community　user　relationship　and　divide　user　group.　On　the　other　hand,　this　method　analyzes　user＇s　interest　which　is　expressed　in　the　published　articles　and　reviews　based　on　support　vector　machine　(SVM)　classification.　We　use　clustering　to　divide　group　based　on　the　characteristics　of　the　user＇s　interest.　A　comparative　analysis　of　the　different　results　can　get　the　final　result　with　higher　accuracy　and　reliability.　At　last,　we　use　software　of　social　network　analysis　to　evaluate　the　results.　This　method　provides　a　theoretical　basis　and　technical　means　for　network　community　application＇s　optimization　and　personalized　services.　©　2013　IEEE.