Possible Algorithms of 2NF and 3NF for DBNorma- A tool for Relational Database Normalization

Abstract—DBNorma [1] is a semi automated database normalization tool, which uses a singly linked list to store a relation and functional dependencies hold on it. This paper describes possible algorithms that can be used to normalize a given relation represented using singly linked list. These algorithms are tested on various relational schemas collected from several research papers/resources and output validated. We have also determined the time required to normalize a given relation in 2NF and 3NF and found that it is proportional to number of attributes and number of functional dependencies present in that relation. The time required, on average, is in order of 186 msec for 2NF and 209 msec for 3NF. By observing time required, one can conclude that these algorithms can be used for normalizing relations within the shorter time; this is specifically needed when database designer is using a universal relation. Index Terms—Relation,functional dependency, normalization, normal forms. I