How to model a business and its data?

As in our previous post, database models refer to the way information is organised and stored. More specifically, models address the way information is related to one another. For eg., in a retail business, we are concerned by the relationship between the customer and the seller, the products sold etc.

Modeling of data requires a thorough understanding of the business. Any business will comprise of a set of workflows involving a set of people and systems performing specific activities. The workflow helps to capture the people-people, system-system and people-system interaction.

A business generally gets initiated by a people-people or people-system interaction. An example of this is a customer visiting a shop asking for a product. Here this is a people-people interaction. Another example would be, a customer visiting a website and ordering for a product. This is a people-system interaction.

Once a business gets initiated, further activities might involve system-system interaction, depending on the business. For eg, in the case of a business involving a customer approaching a stationary shop for an item, it might not require any system-system interaction, but a website selling products online would require a set of systems, like the order management system, payment system, shipping system etc.,
Once we identify the set of workflows, we have to identify the different activities performed in each workflow.

Any business will involve set of activities performed by a set of people and/systems. The starting point of modeling data is to identify the following
1. The different types of people involved.
2. The different activities performed by the people.
3. The data involved in the different activities.

For eg., in the case of a retail business, the customer, retailer, wholesaler, manufacturer are all the different people involved. Each of them have a set of activities or functions. For eg., the customer buys goods from the retailer, the retailer orders goods from the wholesaler and sells goods to the customer. The retailer settles the sales amount with the wholesaler etc. The different types of data that are involved are - the product details, the sales volume and amount etc.

Further drilling down the activities, we capture the individual tasks in each activity/function. Each of these tasks are captured at the granular level as much as possible. Similarly data is also captured at the granular level. For eg., a customer approaching a shopkeeper, asking for a product, the shopkeeper giving the product, the shopkeeper generating a bill, the customer paying the amount, the shopkeeper giving the product and the bill to the customer - all are individual tasks in the particular activity "the customer buys a product".

How is data stored and retrieved from a database?

Data is stored, maintained and retrieved in a database using a set of softwares called Database Management Systems (DBMS). One can compare a DBMS to a shopkeeper. A shopkeeper maintains a track of the goods, piles them in proper places and retrieves the goods when a customer buys.

Similarly, a DBMS does the job of storing information in the appropriate tables, maintaining a track of information stored, retrieves the information when the user gives a query to the database. More than just storing and retrieving information, a DBMS has few other aspects. Following are the different aspects of a DBMS
1. Modeling of data - This refers to the way information is organised and related to each other. There are different methods currently followed, like hierarchical, network, relational, object etc.,
2. Data structure - This refers to the way data is stored, what a record specifies, the fields contained in a record, the field definitions etc.,
3. Query mechanism - This refers to the component that helps maintaining the databases in terms of queries. The various aspects required in maintaining a database are security, user rights and permissions, data storage and retrieval etc.,
4. Transaction mechanism - This refers to the set of activities which are related to each other. This can be compared to the dealings related to a single customer in a shop. The dealings of one customer should not affect the dealings of another customer. Similarly, one transaction should not affect another transaction.

Trivia:
1. Relational Database Management System(RDBMS) is one of the most commonly used Database Management Systems(DBMS).
2. Some of the DBMS products available in market are

Product	DBMS Type
IMS	Hierarchical
Microsoft Access	Relational
Microsoft SQL Server	Relational
IBM DB2	Relational
Informix	Relational
Teradata	Relational
Oracle	Relational

3. One example of a Navigational system is XML. As such there are no specific DBMS softwares exclusively supporting XML. But modern RDBMS softwares like Microsoft SQL Server 2005 support XML using XPath queries.
4. As a RDBMS system, Teradata mainly caters to Datawarehousing applications, unlike other RDBMS systems, which cater mainly Online Transaction applications or a mix of both.

How databases are stored in a computer?

Each piece of information in a computer is stored as file in its hard disk. There are basically two ways of storing information in a hard disk. One, as a text file, which contains only textual data; and another, as a binary file, which contains non-textual data, like pictures, audio, video etc.,

Databases themselves are stored as files in the hard disk. The primary difference between a database and a file is the way information is organised in a database. Typically, information in files would be organised in a way as to reproduce it in a sequential format. Further, information in a file would be something that makes sense as a whole bunch, not in parts. For eg., audio/video files would have information in such a way that when the files are opened using a audio/video software, the information is reproduced sequentially.

In contrary, information in databases is organised in such a way that parts of information could be accessed easily. Further, parts of different information could be accessed and produced in different combination to give different view of the data.

For eg., one view of the Address book would be a list of contacts in a particular city. Another view would be a list of contacts with the first names starting with the letter 'A'. In addition, if the database contains the details of amount lent/borrowed to/from each contact in a Transaction table as shown above, then one might create a view of outstanding amount for each contact in a particular city. In this case, this view is a combination of data from both the Address book with a filter specifying the city and the details of amount lent/borrowed in the Transaction table.

Miscellanea:
Different DBMS use different methods of structuring data. The table structure discussed above is the method used commonly by Relational Database Management Systems (RDBMS).

A simple way to identify if a file is a text file or a binary file is to open in notepad. A text file (like a .txt file) will be human readable, while a binary file will not make any sense to the reader.

A database is usually stored as a binary file. Possible reasons would be
1. Databases may store binary data like images.
2. To store information in an optimized way so that data can be read/written in an optimized way.

What is a database?

Database is a way of organising information and storing it in computer. Information can range from a simple address book to the details of day to day dealings of a big corporation.

Information in databases is usually organised in terms of tables containing records. In the case of a simple address book, a record, which is a row in a table, would contain the information of a particular contact. The columns in the table would contain data for each sub-information within a contact, like Name, Address, City, State etc., for a particular contact.