WITH t AS ( 
  SELECT '1'A,'A'b,'00001'c,'같은내용'd FROM dual 
  UNION SELECT '2','C','00001','같은내용' FROM dual 
  UNION SELECT '1','C','00002','같 은 내 용' FROM dual 
  UNION SELECT '2','B','00002','같 은 내 용' FROM dual 
  UNION SELECT '3','A','00002','같 은 내 용' FROM dual 
  UNION SELECT '1','A','00003','같은 내용' FROM dual 
  UNION SELECT '2','C','00003','같은 내용' FROM dual 
  union select '3','B','00003','같은 내용' from dual 
  ) 
 SELECT C 
    ,MAX(A) KEEP(DENSE_RANK FIRST ORDER BY C DESC) A
    ,MAX(B) KEEP(DENSE_RANK FIRST ORDER BY A DESC) B
    ,MAX(D) KEEP(DENSE_RANK FIRST ORDER BY C DESC) D
  FROM T
  GROUP BY C

-- 1. KEEP 을 이용하는 방법 --
-- d 가 c 에 종속적이라면 Group By c,d
-- 그렇지 않다면 Group By c 하고 d 도 KEEP 을 이용해 구함.
SELECT c
     , MAX(a) a
     , MAX(b) KEEP(DENSE_RANK LAST ORDER BY a) b
     , d
  FROM t
 GROUP BY c, d
 ORDER BY c
;
-- 2. Row_Number 를 이용하는 방법 --
SELECT c, a, b, d
  FROM (SELECT c, a, b, d
             , ROW_NUMBER() OVER(PARTITION BY c ORDER BY a DESC) rn
          FROM t
        )
 WHERE rn = 1
;