How to write a self-join query in hive for avoiding custom duplicates

Question

I have a requirement to get the pair of items by means of the value matched from a table with schema Item, value. I can achieve it by doing a self join but I'm getting duplicate values like depicted below

Item     Value
---------------
item1    value1
item2    value1
item3    value3
item4    value2

When I do self join with distinct, i get values like

Item1    Item2    Value
------------------------
item1    item2    value1
item2    item1    value1

But for me, the above rows are duplicate and need only one of them. How can I achieve this? Appreciate your interest and help.

Note: As I have my own definition of duplicated here in this requirement, i referred it as custom-duplicates in the question. Please do suggest if they are called by a different name.

Answer 1

The solution is simple.
Add in the where clause : Item1>Item2