WITH T1 AS
(SELECT 'Кулон лон слон слон слон Книга книга ' STR
FROM DUAL
UNION ALL
SELECT 'Мама мыла раму раму мыла мама ' STR
FROM DUAL
UNION ALL
SELECT 'ER  ER erer   erer  er  er  erer  ererer  erer  erererer   erererer' STR
FROM DUAL),
T2 AS (SELECT ' (\w+)[ ]+\1' RX FROM DUAL)
SELECT CASE WHEN LEVEL = 1 THEN STR WHEN LEVEL > 1 THEN ' ' END "Символьная строка", LEVEL, REGEXP_SUBSTR(STR, T2.RX, 1, LEVEL, 'i') "Повторяющиеся слова"
FROM T1, T2
WHERE REGEXP_SUBSTR(STR, T2.RX, 1, 1, 'i') IS NOT NULL
CONNECT BY PRIOR STR = STR
AND LENGTH(REGEXP_SUBSTR(STR, T2.RX, 1, LEVEL, 'i')) > 0
AND PRIOR DBMS_RANDOM.VALUE IS NOT NULL;