Biplav Srivastava  Biplav Srivastava photo         

contact information

Distinguished Data Scientist and Master Inventor, ACM Distinguished Scientist and Speaker, IEEE Senior Member


People, known and unknown, have often approached me asking different types of questions related to research, career, technical ecosystem in Indian or just about what life's priorities should be. I do not consider myself as an expert on these topics. However, to the extent that I can see a general pattern and have a view, I plan to share the same on this forum that anyone, but especially the students, could use them at their risk.


Acknowledgments: My views have been shaped by my experiences arising from the discussions and work I have done with my collaborators, teachers and friends. All mistakes are, however, mine.


(c) Biplav Srivastava, Feb 2016

See here for by blogs on technical topics, including Smarter Cities and AI/ Data.

Open Analytics and Transportation (July 2014)

I believe we now have an age of "open innovations" - a confluence of open data (e.g.,, and analytics made possible over the web using open APIs (e.g., Open 311). This has far reaching implications for society.

Here are two triggers which are shaping new possibilities for open innovation in transportation.

1. [Bringing efficiency together to both transportation operators and citizens]

Traditionally, cities treat traffic management as an issue of efficiency for transport operators that they fully or partially fund while citizens are treated as service recipients. Cities had little or no framework to access citizen’s information related to travel and in turn, would pay minimum attention to citizens’ concern of optimizing their own resources. Traffic management is changing with new solutions seeking to optimize both operators and citizens resources together. As example, the cost to run a bus service per km is balanced with what the citizen must pay to travel per km.

2. [From Big to “Big Open Social Data” for Open Analytics]

The role of data in transportation is highly under-appreciated. It was always Big data but ended up being siloed, hard to reuse (due to access reasons) and difficult to do analyze (due to ambiguous semantics). The nature of Big data collected in transportation is changing and new sources of data are becoming available for fusion, like open data from additional government agencies and social data sourced from people. Combined with advancing trends in enterprise integration, analytics and an internet-era collaboration culture, we are heralding an era of open analytics that is easier for citizens, governments and business to consume.

On IBM's 100 years (July 2011)

Free goodies are good, and some were lavished on the 100 years. But one needs to ponder more. Why IBM's 100 age is important is that it has done it in an industry where companies come and go easily -- e.g., DEC. A Google is 10+ years old. So attaining 100 and be still on top of the game is very creditable. Outside IT even, how many Indian companies are 100 years old? How many global companies are this old and doing well. Very few. But it means we, as present employees, have a tremendous responsibility as well. We, from top most managers to lowest rung colleagues, need to focus on universal attributes of good work -- honesty/ethics, top quality of work and being nice to clients and fellow employees. Then only will we see another 100.

Bureaucracy - a litmus test to detect it (16 Sep 2010)

Have you ever felt that an environment has become bureaucratic? An undesired benefit of being in a historically feudal and hierarchy-obsessed society like India, and in a large company, (both unrelated), is that one gets many opportunities to feel it. Here is a working definition to detect bureaucracy. If actions are being taken that beat the common sense rationale, then it is bureaucratic behavior to satisfy a questionable process. As example, assume supply-demand link as common sense as any economics course will teach you. Now, if you have money and there is coffee machine in front of you, but you still cannot buy it, then, it is bureaucracy. You have excess grains that will otherwise rot. You have people who are hungry. And yet people cannot be given grains. That is bureaucracy. Life is too short to serve processes and bureaucracies. Serve people instead. Everyone will benefit.

Max 3 Phrase Rule (circa 2008)

People like to label everything - persons, organizations, countries, etc. Further, we do not remember more than max 3 labels for an entity. Let's do a quick experiment. Think Einstein, Gandhi, China, White House, Romans. How many labels came to mind? Not more than 3!
Given such important entities are remembered by so few labels, it is important to be aware of the labels (which we can also call our legacy). One can either be passive and wait to learn their labels or be active and with their own work, influence what they get labeled with. Our actions are our attempts to design our own labels.

My philosophy on life (circa 2009)

Life is short to be miserable. Do things which excite you. If the current ones don't, move to those that do. If you can't move to new things and still are miserable, learn to love whatever you are stuck with. When you are excited about whatever you do, quality work, money, recognition, and other trappings will follow. At work, follow "Max 3 Phrase Rule" (see above).

My mantra (guidelines) for good research (circa 2008)

  • Understand what is research. Research always sounds jazzy but it is the name of a very disciplined process. What I know and am talking about is called "Scientific Research" as explained here.
  • No matter how complicated the problem area, define a clear-cut problem that you will attack. It may not be the full problem but atleast, the defined problem should have clear inputs and clear expected output. Remember, defining a problem is half the way to solving it. Many people are never able to articulate the problem. Hence, they need to spend a lot of effort communicating that their solution to their perceived problem is valuable enough.
  • Read lots of technical papers and keep written notes. Once you write, you understand what you learnt from related work. The quality of your work will be perceived relative to the related work. Hence, be on top of this.
  • Always think of experiments that can demonstrate the benefit of your solution. You will not always be able to or need to do the experiments but the act of thinking about it will make you take a dispassionate view of your work.
  • Work for and promote a free technical environment. Welcome all ideas and objectively dispose them off. Be free to talk to anyone who can help. Acknowledge all help. Be open to share credits enthusiastically with anyone without whom the solution could not have been put together. Make sure all co-inventors (co-authors, ...) are comfortable with the level of their contributions before they get credit. In case of any conflict, remember that there is plenty of credits to go around. Your objective is to do high quality work and as long that is accomplished, rest will not matter in the long run.
  • Set a quality bar for your work. Once set, follow it religiously. There will be pressure of quantity. In the long run, I believe, quantity is a moot issue (Max 3 phrase rule) and even if you care, both quality and quantity can be accomplished by focusing on quality. Why: because you attract high-quality collaborators who make you more productive.
  • Publish. That is the only way to contribute back to technical field and also give high-quality differentiation to your research sponsors.