E-learning explorations through PhD

It was Leena Ragha of RAIT, Nerul, who got me started into guiding PhD work. It was something that interested me much. Though apprehensive initially, I could get some quality work through my first set of students – Leena, being the first. I set my own standards, and topic preferences for PhD. It was a period when PhD guides in Computer Science in Mumbai was an endangered species, and so many people eventually approached me. Given my commitments at CDAC, and the effort needed for keeping my quality standards, I restricted my numbers to just 2-4 at a time. And for the same reasons, and due to the part-time work nature of many of my students, most of them took about 5-6 years to submit their thesis. Given the desperation of the students to be a doctorate, and that of the institutions to declare higher and higher numbers, the usual norm was 3 years!

I did writeup some aspects of my experiences as a small booklet to help people look at PhD and research in a more serious and meaningful way. Many of the pieces of that booklet were also published on this blog. The main theme was that research is not about PhD or papers, but about developing a mindset – or rather re-acquiring your mindset, which you lost somewhere along your childhood. Many wrote to me saying, the booklet was very helpful. Of course, I have no idea how many of them actually followed any of those guidelines and directions.

I used the PhD route to explore some research directions in technology for enhancing teaching/learning, starting with the language learning work with Abhijit Joshi from DJ Sanghvi College of Engineering. More on this work in a seperate post. It led to some usable structure for building language learning systems. Deepti Reddy took this up further, in the area of student models for language learning. Once she got admission into IIT Bombay, this thread had to be abandoned. She is pursuing a different thread there, in the area of teaching/learning of thinking skills.

Preeti Khanna, at the same time as Abhijit, was assigned emotion recognition as her area. My eyes were on e-learning, looking at the possibility of using automatic emotion recognition to expand the scope of personalisation. If we could sense the comfort level and emotional state of the learner as he goes through the learning process, we can customise instruction, pedagogy and interactivity accordingly. She explored speech, face and keyboard behaviour as sources for emotion. Though, the work did not get to e-learning at all, it gave some insights into the effectiveness of these sources and the then State of the art in this area. This thread is currently idle.

In the next installment, Rekha Ramesh, Zainab Pirani and Rizwana joined as my students. Rizwana wanted cloud computing, and since I was also getting interested in this area, that became her focus. The others were in e-learning related topics. Rekha’s work was in assessment, with the interesting question, “how does one know if a question paper is good?”. Some initial work was done informally with me, and then later she also got into IIT. However, we were able to continue this thread there, with me joining as adjunct professor in the ET dept. I will talk about this work elsewhere in detail.

I accidentally bumped into Zainab Pirani at NMIMS, and she was looking for a guide cum topic. I was getting interested in the field of accessibility. I was CI for a project on making computer desktops accessible for people with various kinds of disabilities. And looking at the challenges of people with mild cognitive impairment in use of computers interested me. Moving further, I discovered the field of Learning Disabilities, and decided to look at what can be done using technology for this. It was also the time, the Anurup project was winding up, and personalisation seemed to be a natural direction to start this new exploration. She built a framework for transforming learning resources for an LD learner. Again we will discuss this in detail, elsewhere.

Amarjeet’s background and interest in AI, prompted me to look at automatic evaluation of subjective answers – a topic that I have been interested in for long. It was (and still is) a hard task, given the variety of questions, the variations in answers, etc. We are trying to look at specific question types, and how we can get an approximate match with a few given sample answers. This is work in progress.

With UGC blocking PhD guidance by people outside the regular faculty of the institute, this whole opportunity has suffered a lot. Most institutions (e.g. NMIMS, etc) are not permitting outside guides.


Tags: ,

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: