As　a　method　built　upon　spectral　graph　theory,　spectral　clustering　has　the　advantages　of　processing　data　with　any　spatial　shapes　and　converging　on　global　optimal　solutions.　But　it　also　suffers　from　the　defect　that　the　clustering　result　is　quite　sensitive　to　its　parameter.　A　minor　change　of　the　value　of　the　parameter　affects　the　clustering　accuracy　greatly.　In　this　paper,　a　novel　approach　which　integrates　grey　relational　analysis　based　on　difference　information　theory　with　spectral　clustering　is　proposed.　The　similarities　between　data　points　are　described　by　the　balanced　closeness　degrees　of　their　attribute　sequences,　so　that　the　impact　of　the　parameter　is　eliminated　and　the　performance　can　be　improved　simultaneously.　The　experimental　results　proved　the　effectiveness　of　the　new　method.