Date: 01-17-2022

We all know that databases were created to deal with data and its storage. We are all confused about which database to use as we have lots of options to pick from! Generally, we choose the database producer or the owner. We can select the right database for our needs by analyzing the different types such as hierarchical and relational. There are other types such as network databases and object-oriented databases, but we won't be covering them here.

What is a Hierarchical database?

In a hierarchical database, the data is organized in a tree-like structure. Each individual data is stored in a field and the fields, in turn, form records. These data are accessed with the help of links between them. In this structure, all the data records are linked finally to a single parent record. It is also called as the owner record. The links between the records are often described as parent-child relations. The best usage of a hierarchical database is its deployment in a library system as it stores names or book numbers using the Dewey Decimal System. This system resembles a tree-structure by sharing the same parent number and then branches like trees. Similarly, we can use it to store names in a phone directory.

What is a Relational database?

A relational database stores data in the form of tables with unique keys to access the data. These tables supply the data in the required form with the help of query languages. The interesting part is that it does not require any data re-grouping to fetch the data of choice. It is often referred as a relational database management system (RDBMS).

Differences

• Simpler to use: The hierarchical databases use the logical parent-child relationship and it looks simpler as well. But relational databases involve tables to store records in the form of table fields. Also in most cases, it requires a unique key for each record.
• Which is older? The hierarchical databases came into existence before the relational databases and, it is the precursor to all other databases.
• Fundamental difference in the data notion: In hierarchical databases, a category of data is termed as 'segments' whereas in relational databases it is termed as 'fields'.
• Inheritance: Every child segment/node in a hierarchical database inherits the properties of its parent. But in relational databases, there is no concept of inheritance as there are no levels of data.
• Data linking: In hierarchical databases, the segments are implicitly linked as a child is linked to its parent. But in relational databases, we should explicitly link the tables with the help of 'primary keys' and 'foreign keys'.
• Use of keys: The relational databases are generally framed with unique keys called primary keys and also keys from other tables called foreign keys. These foreign keys are primary keys in some other table, and it is referred to while accessing the other table from this table. The main use of keys is to give a unique identification to the data records and to refer other tables during the data fetching process. But a hierarchical database never uses keys. It has its links to denote the path to be traversed during the data fetch. Therefore, we can consider the keys in relational databases as the equivalent of the paths in hierarchical databases during data fetches. The paths do not represent the uniqueness of data that has been stored in hierarchical databases.
• Unique and duplicate data: As the keys represent the uniqueness of data in relational databases, we can easily list such data on demand. When the same is required in a hierarchical database, it needs a lots processing. We can have more than one copy of the same book in a library but assigned with different book numbers. In this case, we should compare the book names to identify the duplicates. Therefore, relational databases are suited to store unique data whereas hierarchical databases are good for data with duplicates.
• Data fetching: Just imagine that you have a library management system, and it stores book details with an assigned book number for each book. Consider a book assigned with the book number as 1034. The data fetching process here is just given below:

• In a hierarchical database:

If book-no > 500 {... } If book-no > 1000 { If book-no >1500 { ... }

Else {if book-no >1100

If book-no > 1050 {...}

Else {if book-no >1025 {if book-no >1030 {if book-no > 1035 { ...}

Else {if book-no = 1031} ...

Else {If book-no = 1032} ...

Else {If book-no = 1033} ...

Else {If book-no = 1034} ... Match found here

The above process takes place step-by-step as we reach a branch of the tree climbing from its trunk.

Consider that we need to fetch the 'date of birth' field whose employee-ID is 12345. Here the employee-ID is the primary key and we frame queries as below:

• In a Relational Database: Here, the data is fetched with the help of the Primary keys and the Foreign keys. We can directly access the required fields with its matching key.

Fetch Employee-name, Employee-DOB

From Employee-table

Where employee-ID = '12345'.

Here we can fetch the required fields directly and we need not beat about the bush.

• Many-to-many or one-to-many data linking: These kinds of data links are not possible with the hierarchical databases as a parent can have more than 1 child whereas a child cannot have more than 1 parent. In the latter case, we would encounter the many-to-one or the many-to-many data linking or relationship. But these kinds of data relationships are possible with relational databases.
• The fields in relational database vs. the nodes in hierarchical database: In relational databases, the data classification is based on the 'field' whereas in hierarchical databases it is based on the 'nodes or segments'. Every field is present in every record in relational databases. Similarly, we can see every segment in the final data i.e. book number, book name, etc. in the case of a library management system. This is often referred as the fundamental difference between the two databases, which we have mentioned at the initial stages of our article.
• Where does it find its usage? Each database finds its usage in an application or system and is purely based on the requirement. For example, the library management systems use a decimal system that numbers the books similar to a tree. In these systems, the RDBMS does not work well as its concept is different. But when we consider an organization, the details of employees or goods cannot fit a tree-like structure. Therefore, tables can be a better solution to store such details. So, here relational database is a better choice.

Comparison Chart