if its the usually unf to 3nf its very easy but my school have their own method and if its difficult I will be glad but its impossible some times.
this are the rules..
Un-normalised Form – UNF: Transfer all identifiable attributes from the sample to a list ensuring each attribute has a unique name; structure the list according to the repeating groups (a repeating group is defined as any group of one or more attributes which has [regular sets of] multiple values for a single value of the initial key); choose an initial key for the RDA
First Normal Form – 1NF: Separate into new relations each repeating group, generating appropriate keys for each relation: this means choosing a key for the group and propagating (copying) the key of the relation from which the group was separated as part of the key of the new relation; any single-valued attributes remain with the initial key
Second Normal Form – 2NF: Separate into new relations any attributes dependent on only part of each compound key along with their non-key attribute(s); the original compound keys are left intact (along with any dependent attributes)
Third Normal Form – 3NF:
Separate into new relations any attributes dependent on other non-key attributes along with their key attribute(s); foreign keys are retained in the original relation
Check compound keys for any redundant parts – if any part of a compound key can be determined by the rest of the key, demote the attribute(s) to non-key status (these may also be foreign keys)
...could be easy in theory