From my understanding through the lecture and slides, "UNIQUE" is specified when a Child occurrence is associated with just one Parent. Does it indicate that Child and Parent have one to one relationship when both of their cardinalities are 1 regardless of optionalties (i.e., 0..1, 1..1)? If not, can you give me further details when to add "UNIQUE" in logical model?

I'm reviewing the slides with four examples above, but a bit confused about it. I'd appreciate your help. Thank you.

at the first two table, why name use 'constraint name unique not null', why it not in primary key?

Could you explain please how we logically model an ER diagram where there is only one entity type which has a relationship with itself.

This was the correct answer in the quiz:

CREATE TABLE DOG_WALKER (DOGWALKER_ID, FIRST_NAME, SURNAME)

PRIMARY KEY (DOGWALKER_ID)

CREATE TABLE FRIENDS (DOGWALKER_1_ID, DOGWALKER_2_ID)

PRIMARY KEY (DOGWALKER_1_ID, DOGWALKER_2_ID)

FOREIGN_KEY  DOGWALKER_1_ID REFERENCES DOG_WALKER(DOGWALKER_ID)

FOREIGN_KEY  DOGWALKER_2_ID REFERENCES DOG_WALKER(DOGWALKER_ID

I don't understand why and how we would split DOGWALKER_ID into 1 and 2?

Thank you

Is it true that for generalisation and composition relationships these can be modelled as one-to-one relationships? Will they always have 1..1? Also, do we always know the cardinality when reading from the other direction, so for specialisation and aggregation relationships?

Thank you