Prof. Animesh Mukherjee

Next in the series of Know Your Professor(KYP), is the young Dr Animesh Mukherjee, Associate Professor, Department of Computer Science and Engineering who has been doing pioneering research in the field of Social Media Analysis and Information Retrieval and has made fundamental contributions to the field of computational social science. Recently he was selected for the ‘Ethics in Artificial Intelligence (AI)’ Research Award, winning the proposal for Machine Learning Education with Tensorflow 2.0 in addition to 3 research track papers getting accepted in CoDS-COMAD 2020

Also, having experienced KGP as a student while pursuing his Master’s and PhD degree, he has essentially been on both sides of academia in KGP for more than a decade and has witnessed KGP evolve and ascend into an institution of global stature.

You have a lot of passion for research. What inspired you to pursue a career in research? What is your source of motivation for research in the field of Language Dynamics, Social Media Analysis and Information Retrieval?

I joined as an M. Tech student here at IIT Kharagpur. In the second year, I got a project with Prof. Anupam Basu, who was administering the Communication Empowerment Laboratory. The laboratory was developing tools for people who have motor problems like cerebral palsy. Together with Dr. Basu, I developed a software mouse. I got connected to the Indian Institute of Cerebral Palsy, where we had an opportunity to deploy our software and see how people felt. This developed an interest in me for pursuing research. During those days, I came across Dr. Monojit Choudhury. We used to discuss a lot about the intricacies of human languages — how it changes over time and place. He was one of the prime inspirations for me to join the research field. My thesis work during M.Tech gained popularity, and I got an award for it.

While I was doing my Post-Doc, we were working with models, mostly mathematical. We imagined the behaviors of people in a linguistic society and approximated it using our models. Then we saw that social media could be our way to study languages. Social Media is a platform through which users evolve their language styles and allows people to time-travel as social media time-stands all your posts. Thus, the same interest drove me to Social Media Analysis.

What were the hardships that you faced during the journey of becoming a professor?

There were some social hardships, but fortunately, I did not face many technical difficulties. While doing my Post-Doc, I wanted to continue it for one more year. However, my father got fatally ill, so I had to give interviews at different IITs hastily. I was sure of coming back to academia as I believed that teaching to be the right profession for me.

One of the hardships that I think still bothers me sometimes is joining the same department from where I graduated. The problem was that suddenly one day, all of my teachers became my colleagues. It was challenging to absorb this change and put it in my psyche. In many cases, I would feel to express something, but then I would feel it might be inhibitive as I saw my teachers in front and not colleagues. It took around 2–3 years to get away with this perception.

Another hardship is that when you come back to your group, making an individual identity becomes difficult. You would be known as the Ph. D. student of your guide when you join. So, to make your own identity, it takes much more effort. Everybody will know that these professors led this group, and this guy is another fruit of this tree. So, to become a tree in yourself takes a lot more effort.

What, according to you, are the differences between students at your time and current time (in terms of the education system, difficulties faced by students, career goals, and passion towards engineering)?

In my time, there were comparatively fewer students in any discipline than there are now and a considerable fraction of people went for higher studies, at least masters. This, somehow, has depleted in the last few years. People are less interested in going for higher education. The reason may be that companies, these days, pay very well and hence, students do not feel the urge to study further. The issue can also be with the teachers; maybe we are not able to motivate students well. The main problem that comes is that in a class with forty students, groups of 4–5 people could be given a term project, which is very easily manageable. But the moment the strength becomes 180, the explosion in the number of students reduces the one-to-one connection among students. Some of them might not even be motivated.

Also, I believe that students should be decent in class and should ask more questions. In many of the courses I teach, just 10% of the students ask doubts, which means that either the rest are not interested in the class, or they feel that they would end up asking a wrong question. One should understand that there is no harm in asking the wrong question.

You did your Masters and Ph.D. in IIT Kharagpur. What difference do you see here from the perspective of a student and a professor?

I had much more flexibility as a student. Students have to worry twice a semester, but I have to face them daily and teach something. So, it is like never-ending examinations for me. As a student, there was the pressure of Ph.D. papers, deadlines, etc., but I had to manage myself only. Now, as a professor, I have many lives to manage: the group of students whose classes I am taking, the group of students for whom I have been faculty advisor, and the group of students who are pursuing their Ph.D. with me.

Apart from academic issues, there are administrative issues, family issues, etc. Hence, things have become much more complicated than the student lifestyle. However, there are some vent outs. Teaching in the classes and doing maths with my little daughter are two things that recharge me a lot.

You were a junior researcher in ISI Foundation, Italy. How was your experience there? What difference did you observe in students and teaching lifestyle there compared to India?

It was an exciting experience, both the social as well as the academic environment. I was made aware that I was a Post-Doc researcher then. Gone were the days that people would guide me. I had to guide people now, and my advisor would come once a month. There, I had to develop questions on my own. If somebody asks a question to you, then finding an answer is not so difficult. Instead, the harder part is to find a good question to answer.

Students in Turin are much more independent. They would perform judicial scrutiny before every new course that they would take. If they give time to a class, they will extract the maximum from it. In our case, while some students make this decision very thoughtfully, many of them do not. I also try to teach this habit to my Ph.D. students.

What are the hobbies that you like to pursue in your free time?

I like to listen to classical music and watch movies. I usually like movies which are relaxing and entertaining. My sole expectation from a film is entertainment, so if I am not attracted by movies like inception, matrix, etc., where I have to go and try to understand them. I would be much more interested in light-mood movies, which are a source of recreation. The other thing that I like is reading storybooks both in English and Bengali. These days, I am rejuvenating this hobby again by reading storybooks to my kid. Every day, I try to find some time to read a story to her. In general, I enjoy reading a lot.

You were recently awarded the Ethics and AI award. Would you like to throw some light on your research and plans in the domain of Ethics and AI, considering that it is one of the newly emerging fields?

The entire thing is motivated by social media-related research. What I have observed is that social media has become a colony of micro-abuse, and the issue is that here, the impact is more long term. It seems like these platforms have an ethical and moral responsibility. Even if it is not able to distribute profit, it should not distribute harm among the consumers. In many western countries, people decide whether a criminal is going to re-offend based on ML algorithms. It turns out that these algorithms are highly biased. So this proposal tries to address some of the problems in this area.

Presently, we are looking into certain newspapers and news channels and crawl videos, make transcripts from those videos as well as look at the comments made by people. Then, we would see if there is a similar sort of bias present there or not. Another thing that we are planning to do is to understand how these news items are getting shared in social media.

What message would you like to give to the students of IIT Kharagpur?

I have been a student, and I have always enjoyed being among students. The only message that I would like to give in the current world scenario is since you are budding adults, start making informed decisions and not influenced ones. If every decision in life is an informed one, it will produce much better results for you in the long term. It would be much more helpful than those taken under the influence of others