Normalization is used to measure the similarity between 2 trees.

• Max: Using this normalization the edit distance is divided by the number of operations needed to create the largest of the 2 trees from scratch (in terms of nodes). This normalization does not bound the results within [0,1]..
• Sum: This normalization divides the edit distance with the sum of operations to create both trees from scratch. This ensures that any result will be within [0, 1].
• None: The flat edit distance between 2 trees will be returned.