Big Data in practice - Bernard Marr - Review


Bernard’s Big Data in practice was launched in March 2016, the crux of the narrative is to highlight the importance of data in just about every industry. At the start of the book, he gave a good reference point on predicting the amount of data which is likely to be created every day across the globe by 2020, which happens to be 170 GB per day, the actual numbers in 2020 worked out to be 1 billion GB per day, he has missed the mark by a factor of 5.8 million times. It goes to show the explosion of data in the real world and its importance is foolhardy to undermine. 

Bernard Marr has assessed Big Data implementation and use cases in 45 different companies, a few of them happen to be truly path breaking. In the aftermath of 9/11 terrorist attack on WTC twin tower, the US security agency had the added responsibility of identifying bad people from 7 million visitors who checked into the US about a decade back. Irrespective of the number of highly trained folks deployed in the airport security checks, the task is daunting beyond measure. AVATAR (Automated Virtual Agent for Truth Assessments in Real-Time) had been deployed by the US government to keep a check on millions of visitors. 


The AVATAR relies on three sensors to make probabilistic judgement on the truth of statements given by the travellers. An infrared camera data on eye movement along with pupil dilation at 250 frames per second. A human undergoes higher cognitive load during deceptive response which results in increased pupil diameter, termed as dilation. The second sensor captures the body language and flags suspicious movements, while the third one is used to capture noticeable changes in tone and pitch of voice. Using inputs from all the three sensors in real time, AVATAR flags suspicious travellers and provides scores on various parameters of bad human behaviour. The system continuously learns by itself using the mountains of data points, arriving at a cutting edge solution. The number of terrorist incidents post the disaster is for us to see and AVATAR certainly has a telling contribution, if Big Data can be reliably applied on such critical functions such as airport security, the use cases are bound to be omnipresent. 


Earthquakes have historically proven to be not just difficult but impossible to predict, as called out by seismologist. It’s astonishing to note that such a complex occurrence can be predicted using Satellite Big Data and Terra Seismic has been pioneers in the space of earthquake prediction with 90% accuracy across the globe. Satellite images are captured, atmospheric sensor are used to understand various conditions and even unusual cloud formations are used as variables to create earthquake prediction model. It’s a telltale sign on the undeniable penetration of Big Data and AI into our livelihood. 


We are now in the year 2024, Bernard would have commenced drafting his book at least a decade back, purely assessing the duration from Big Data velocity, the book is a century old but it's eye opening to understand the depth of use cases cutting across 45 varied industries and hence worth a read. 


Bernard Marr has over a million followers in linkedin and is recognized as a respectable voice in the data business. Big Data enthusiast can follow his podcast in https://bernardmarr.com/podcasts/

Comments

Popular posts from this blog

CFA preparation guidelines for level 1

Review : Imagining India

The trip to Delhi (5th - 11th Aug 2007)