Semantic　relations　acquisition　is　a　crucial　work　in　the　field　of　knowledge　acquisition.　This　paper　presents　a　method　that　acquires　semantic　relation　patterns　from　large　microblog　text.　It　initially　analyzes　the　characteristic　of　microblog　text,　and　give　an　algorithm　of　acquire　patterns　semi-automatically.　Semantic　relations　are　extracted　from　microblog　corpus　based　on　concept　recognition　and　pattern　matching.　The　experimental　results　show　the　performance　of　extracting　semantic　relation.　©　2014　IEEE.