Data Science is not covered in a single step. It follows a series of steps that are helpful in solving a problem. This is what we call Data Science Life Cycle (DSLC). DSLC usually comprises of the following steps:
Step 1: Understanding the Problem
Step 2: Collecting Data
Step 3: Cleaning and Validating Data
Step 4: Analyzing Data
Step 5: Interpreting Data
Step 6: Visualizing and Communicating Results
5.2.1. Understanding the Problem
Before we start talking about collecting data, it is necessary to clearly understand the problem under consideration. Understanding the problem is the most important step in DSLC. In case, we do not understand the problem properly, we may collect the wrong data resulting in useless or wrong results. In this initial step, we need to ask questions (what, why, who) such as:
- What are we trying to find out?
- Why is this problem important?
- Who will use the results?
Identifying and Defining a Problem
First, we need to identify a problem that we want to solve using data. The problem can be related to our school, home, city, or any area of life where decisions are made based on information. Once a problem is identified, then we must define it clearly and simply.
Example: Instead of saying, "students are not happy," we can define the problem in a better way as "Which school activities do students enjoy the most?"
When defining a problem, we need to consider the following: (i) Use clear and simple words, (ii) focus on one specific question, and (iii) avoid unclear or general terms.