In　the　existing　Chinese　ontology　learning,　most　research　methods　on　the　non-taxonomic　relation　extraction　can　not　make　clear　the　concrete　name　of　non-taxonomic　relation.　In　other　words,　they　can　not　attach　a　semantic　label　for　the　extracted　non-taxonomic　relation.　The　paper　presents　a　method　firstly　extracting　correlated　concepts　by　use　of　statistical　function　and　then　getting　the　name　of　non-taxonomic　relation　through　linguistic　representation　which　fills　the　gap　of　most　other　research　methods.　Empirical　validation　shows　that　the　method　proposed　can　help　obtain　the　name　of　the　non-taxonomic　relation　between　concepts.