不需要自联接表;使用LEAD解析函数更有效：

SELECT userid,
       datetime as LoginDateTime,
       LogoutDateTime
FROM   (
  SELECT userid,
         datetime,
         activity,
         LEAD(CASE activity WHEN 'Logout' THEN datetime END) OVER (
           PARTITION BY userid
           ORDER BY datetime
         ) AS logoutdatetime
  FROM   useractivity u
)
WHERE  activity = 'Login';

其中，对于样本数据：

CREATE TABLE useractivity (
  Id       NUMBER,
  UserID   VARCHAR2(20),
  DateTime DATE,
  Activity VARCHAR2(6)
);

INSERT INTO useractivity (id, userid, datetime, activity)
  SELECT 123, 'abc', TIMESTAMP '2023-08-24 09:30:50', 'Login'  FROM DUAL UNION ALL
  SELECT 125, 'abc', TIMESTAMP '2023-08-24 09:34:23', 'Login'  FROM DUAL UNION ALL
  SELECT 150, 'abc', TIMESTAMP '2023-08-24 09:38:04', 'Logout' FROM DUAL UNION ALL
  SELECT 178, 'abc', TIMESTAMP '2023-08-24 13:23:12', 'Login'  FROM DUAL UNION ALL
  SELECT 200, 'abc', TIMESTAMP '2023-08-24 15:50:34', 'Logout' FROM DUAL UNION ALL
  SELECT 300, 'abc', TIMESTAMP '2023-08-24 16:43:02', 'Login'  FROM DUAL UNION ALL
  SELECT 505, 'abc', TIMESTAMP '2023-08-25 10:32:12', 'Login'  FROM DUAL UNION ALL
  SELECT 507, 'abc', TIMESTAMP '2023-08-25 12:36:44', 'Logout' FROM DUAL;

输出：
| userId|登录时间|登录日期时间|
| --|--|--|
| ABC| 2023-08-24 09：30：50| * 空 *|
| ABC| 2023-08-24 09：34：23| 2023-08-24 09：38：04|
| ABC| 2023-08-24 13：23：12| 2023-08-24 15：50：34|
| ABC| 2023-08-24 16：43：02| * 空 *|
| ABC| 2023-08-25 10：32：12| 2023-08-25 12：36：44|
fiddle

SELECT
    L.UserID,
    L.DateTime AS LoginDateTime,
    MAX(R.DateTime) AS LogoutDateTime
FROM
    UserActivity L
LEFT JOIN
    UserActivity R ON L.UserID = R.UserID
                    AND R.Activity = 'Logout'
                    AND R.DateTime > L.DateTime
WHERE
    L.UserID = 'abc'
    AND L.Activity = 'Login'
GROUP BY
    L.UserID, L.DateTime
ORDER BY
    L.UserID, L.DateTime;