Data Science: An Effort to Extract Some Insights from Data

There is a proverb, “Tomorrow never comes”, but in the perspectives of modern computing, more closely with respect to Data Science, it seems little incorrect. It can be said that Tomorrow has come and it is today only and it is creating a new Tomorrow with help of Yesterday.

Modern computing makes an effort to mimic most of the human activities in machines only. In this regard it sees everything as data exclusively. It perceives yesterday and today as data, analyses it, and predicts for Tomorrow.

Earlier data would be processed with the disciplines of Statistics. But due to lack or very less computing power, it was not that much effective. So Data Science came into the existence with high computing power blended with Statistical techniques.

Let’s understand the application of data science with some contemporary projects for which it is being implemented.

• Data science techniques are being implemented for finding customer behaviour, sales patterns, customer requirements, etc., and based on these circumstances recommender systems are being generated for both salesperson and customers.
• Logistic companies such as FedEx, DHL, etc. track down the best duration and route for the shipments for delivering on time with the best mode of transport.
• Forecasting employees’ attrition and to get the bottom of the attributes that may influence the turnover for the employees.
• Developing advanced models using Artificial Intelligence and machine learning techniques which once set in the motion and perform longer with no human intervention.
• Along with genomic data, data science can put forward a thoroughgoing understanding of genetic impediments about specific drugs and diseases.
• Aviation companies can now successfully predict the flight setbacks and intimate the passengers.
• Data from social media are being analysed in the different perspectives and so much of insights are being evaporated.
• ...in this way, there are so many applications which are incorporated with data science.

Let’s now put some lights on the origin of data science.

Before data science, the term data mining was coined and had got publicized in 1996 form the research paper entitled, “From Data Mining to Knowledge Discovery in databases.”
Later in 2001, William S. Cleveland had put forward the term data mining to a new horizon by blending computer science and data mining. Merely, he had incorporated statistics with computing power and called this composition as “Data Science”.

Nearly this time only the Internet had also emerged with the new version “Web 2.0”, assimilating social media welfares such as MySpace in 2003, Facebook in 2004, YouTube in 2005, Twitter in 2006, WhatsApp in 2009 etc., where people also became huge stakeholders for generating huge data by creating texts, uploading images, audio and videos, like and share, and leaving their footprints in the digital landscape. These undertakings were to sow the seeds of a pile of data from so much to too much and beyond, which was termed as “Big Data” that came across a world of opportunities in discovering insights using the data.

So, the rising of Big Data in 2010, forged ahead the existence of Data Science to throw its weight behind the need of the business to extract insights from their massive unstructured datasets. The Journal of Data Science narrates the data science as, “... almost everything that has something to do with data: collecting, analysing, modelling, yet the most important part of its applications ... all sorts of applications”, like machine learning, deep learning, etc.

In the year 2010, with the new horizons of the data, it became a trend to train a machine learning model with the approach of data orientation rather a knowledge orientation approach. Machine Learning and Artificial Intelligence pushed around the media transcending every other features of data science such as exploratory data Analysis, experimentation, etc., and as a whole termed as Business Intelligence. Some more terms and understandings can be outlined from this adjacent figure.

IT+IT=IT, that is Information Technology + Indian(/International) Talent = India(/International) Tomorrow) is a famous quote in the computer world few years back, but now its redefined as DS+DS=DS that is Data Science + Digital System = Developed Society

Let us now see the application of data science to solve a business problem.

Suppose Ms. Leena works for a multinational company as a data analyst and she has been assigned a business problem. So she would be considering the following steps for solving it.

1. Defining the Business Problem or Understanding
2. Data Acquisition or Data Mining
3. Data preparation or Data cleaning or Data pre-processing
4. Data Exploration or Exploratory Data Analysis (EDA)
5. Feature Extraction or Feature Engineering
6. Data Modelling or Predictive Modelling
7. Data Visualization and Communication

Following figure can explain these steps in some extent.

Figure Source: Internet (https://www.omnisci.com/learn/data-science

There may be different job titles in the aspect of data science such as:

• Data Scientist
• Business Intelligent(BI) Analyst
• Data Analyst
• DBA or Data Engineer
• Machine Learning Engineer
• Deep Learning Engineer
• Cloud Architect or Data Centre Engineer
• ... and many more.

Future with Data Science
This is the era of the Data Science which needs only data as input to create magic. So many insights it can extract from the data which may be quite astonishing. Experts say the scope of the data science is eternal and there are a lot more to appear in these perspectives. New rules need to be set, new algorithms and more advanced computing languages are aligned along with more advanced computing power.

So there is a big rival among human and computer also, if we don’t update ourselves parallel or ahead of computers, we may lose our jobs and it’s also utterly possible that we may be kept on a tight rein by machines only.