Question

Suppose you have a file of data of approximately 100,000 personnel records (FYI, 213 = 8192 and 214 = 16384) using Social Security numbers as primary key. You expect to mostly access individual records so you want that to be fast, but you expect to also access the entire file in sorted order reasonably often. You have decided to use an indexed sequential file structure or an order 5 B-Tree. Give 3 reasons why an Indexed sequential search structure is better. Then give three reasons that an order 5 B-tree is better. What additional information do you need to choose between them?

Answer 1

Answer:

See explaination

Explanation:

The reasons why indexed sequential search structure is better are:

1. In index sequential search any field of the records can be used as the key. This key field can be numerical or alphanumerical.

2. Since each record has its data block address, searching for a record in larger database is easy and quick. There is no extra effort to search records. But proper primary key has to be selected to make efficient.

3. This method gives flexibility of using any column as key field and index will be generated based on that. In addition to the primary key and its index, we can have index generated for other fields too. Hence searching becomes more efficient, if there is search based on columns other than primary key.

The reasons why 5 B-tree is better:

1.The B-tree Provides support for range of queries in an efficient manner and You can iterate over an ordered list of elements.

2. B-Tree algorithms are good for accessing pages (or blocks) of stored information which are then copied into main memory for processing. In the worst case, they are designed to do dynamic set operations in O(lg n) time because of their high "branching factor" (think hundreds or thousands of keys off of any node). It is this branching factor that makes B-Trees so efficient for block storage/retrieval, since a large branching factor greatly reduces the height of the tree and thus the number of disk accesses needed to find any key.

3. It is a generalization of a BST in that a node can have more than two children. These are self-balancing and hence the average and worst complexities is logarithmic. We opt for these when the data is too huge to fit in main memory. These structures are used in database indexing and help in faster operations on disk

the additional informations is we should first decide to choose which structure is suiatable for which algoritm in terms of space and time and then use the appropriate search algorithm.