E-learning explorations through PhD

December 1, 2015

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.


Ebasta: Schools books to e-books

November 27, 2015

Ebasta is a project being implemented by CDAC — Mumbai and Chennai centres. Ebasta is a combination of 3 ideas. Firstly, it provides a web portal linking the publishers, the teachers and the students on a single space. Publishers, well, share their e-resources on the portal (http://www.ebasta.in). Teachers can browse all these e-resources, and prepare a bundle of resources for their wards. They can take not only textbooks, but also reference materials, supplementary resources, and even story books. Students can then download the prepared bundle.

Screenshot from 2015-11-27 21:01:09A bundle of such a nature spanning almost a dozen subjects in a school environment, means a large number of electronic files. On a laptop or a tablet that the student may be using, this is a recipe for confusion and frustration. When these are mixed with myriad other files that he/she may download, the situation is quite bad. Also remember that, the websites name their download files for their convenience. For example, NCERT files are named using a coded naming convention. Other sites use their own codes and conventions. This complicates locating of a file into a hard task.

This is where the second idea comes. We introduce the concept of an e-basta — a structured, virtual bag. The bundle can be organised into a hierarchy, like having pockets/compartments in your bag. The teacher can create a bag with as many compartments, and sub-compartments, as he/she likes. And finally the actual books or e-resources can be inserted into these pockets. This puts all the related items together, and are placed at a fixed place relative to the bag. The third pocket contains maths resources, for example. It is this kind of structured bag, the ebastas, that the portal offers to the student for downloading.

Closely related to this context is the third idea. An app on the tablet device who understands this concept. This saves the student from dealing with any files directly, let alone their names and extensions. The app understands ebastas on the tablet, sets them up suitably, and shows very intuitive names for the various folders and resources — these names are as defined by the teacher creating the bastas, and not the publishers or writers. Once an ebasta is downloaded to the device, the app takes care of everything.

A key driving principle behind ebasta is that difficult to get sustained Internet connection for most of the school going students. Therefore, along with the ebasta download, we set up the whole system, so that, no further Internet connection is needed for reading/referring to any of the resources. The publishers relying on digital rights management to protect their resources from piracy is advised to use only device specific and time specific information for this purpose, so that Internet access is not needed when reading books.

We have most of the State Boards slowly coming on board, and even publishers have expressed interest in coming on board — some are already on. So, we will be having a wide collection of authentic e-resources, relevant to school syllabus in the portal. What are you waiting for? Get your resources on the portal — if you are a publisher, or a Board. Browse the resources and ebastas and download whatever you like, if you are a parent, or a student. If there are priced books, remember you will be required to make payment, before the download is given.

Like the work, the concept, etc? No? Tell us why, either way? And share your thoughts and suggestions also.

TITLE — Technology for Improving Teaching Learning Experience…

May 21, 2014

I am making some effort to consolidate the various activities in India in the area of e-learning. This is primarily driven by my own interest in using the various technologies to help the quality of education. In March, we organised a symposium on the same topic in collaboration with JDBIMS, SNDT, which saw about a hundred teachers and practioners join. T4E conference series has been active for a few years. There are active research groups in various aspects of TLP around the country. I am opening a new blog on this topic, where we can discuss this in detail. But for now, here are some titbits which may be of interest.

1. Along with a few likeminded people volunteering to chip in some time and energy, we have started a news letter, named TITLE, which will come out, primarily in electronic form, every month. If you are interested in receiving a sample copy or regularly, drop me a mail or comment on this post. The first issue has been released a few days back.

2. A proposal is pending with CSI — computer society of India — for forming a special interest group in this area. It may happen formally soon. We look forward to your support and enthusiasm in organising events, compiling resources, etc under this umbrella.

– Sasi

Hyderabad visit….

June 30, 2013

Been to Hyderabad during 28-29 June for the Eleltech conference. Gave an invited talk on personalisation. Also visited the upcoming new building of CDAC Hyderabad, and saw some interesting “cloud computing” happening on top of the building. Here are some snaps.

Inauguration of Eletech-2013

Cloud computing in Hyderabad!

CDAC Hyderabad building getting ready....

The Journey of Life…

June 22, 2013

 The road ahead seems never to end;

Out of the blue, comes another bend;

Lovely and smooth on many a part;

At times we are like on a bullock cart;


The journey of life is same indeed;

Potholes on the road – sorrows they seed;

On stretches uphill, it is a heavy load;

On the way down, with joy we nod;


Lovely blooms, meadows and hills;

Just like friends, they bring up thrills;

Narrow stretches, so much flawed;

Testing patience and faith in God;


Sharp bends ring a warning bell;

We may yet know, the route so well;

Moments bringing death so close;

On and on still the journey goes;


The journey of life no different to-day;

Always alert good drivers would stay;


Choosing a course….

June 1, 2013

What course to do next is a question many people face, often multiple times in their life: after 10th, after 12th, after degree, for their children, and so on. Options may be degree courses, short term proficiency enhancement courses, diploma programs and so on. So the domain is very broad and complex. I am restricting, here, to the post-graduate diploma programs that people usually take to enhance their career prospects. CDAC also offers a lot of such courses, and one can witness amusing discussions on the way this problem is addressed in discussion forums and other such avenues. And that incidentally is one main reason for me to make this post! This post is, therefore, an attempt to put together some thoughts which may help some of those asking such questions, and may be of help to even those providing half-baked or biased responses to such questions.

Of course, placement is a major concern in making the selection. No point in denying that placement is why most students go for such courses. But what does placement mean? Getting “a” job? Any job? Any company? And long term or short term? Often these questions are rarely addressed. And wrong decisions are made at a high cost. In a market driven system, no one can guarantee placement. Institutions can make an attempt to get companies to campus – depending course, market dynamics, the institution, etc you get different subset of companies to come. And even then, getting job is dependent on you too. If you are not prepared, nothing is going to help you. Often apart from the interview performance and your proofs of relevant domain knowledge, your background (degree taken, its score, performance in prior education, etc) also may play a role. In short, there is no magic wand to get you a job. Claims guaranteeing a job is like Complan guaranteeing extra-height! And also find out what kind of job people get and on what terms. Some call even trainee-ship as “placement”! So check details before believing anyone.

Not all courses are equivalent. Choosing a course on embedded systems just because it may be offering a higher placement chance is a bad idea, if your interest is not in that area. Look at the syllabus and see if that interests you. Even within computer science, there are different focus areas and profiles. Find out what is the focus of the course, and see if that is what you want. And if yes, look at the following aspects further. And with all that together, your choices will be simpler. Most computer science graduates look for “developer” jobs. Very rarely does application developers write Java/C code from scratch today. So many libraries and development frameworks are available, for many applications, you need hardly any code to be written. These are also called “developer”; nothing wrong in that because you are “developing” an application. But it may not match your image of a “developer” — which may be writing full programs in a language like C/C++/Java/etc. Writing some raw code is not the most challenging part of computer science. Figuring what code to write and how to organise the code is. Designing the architecture of the software, the various classes/modules/components, the data storage and representation, the test scenarios for them, etc are the critical areas today. Dont fall for the “developer” image without thinking…

In the area of IT/CS, practical skills are important. Our universities generally do a poor job of this aspect, for various reasons. Most of the time, the culprit is not the syllabus, but its implementation! And lack of practical skills is the major bottleneck people find in getting jobs, and various diploma and certification programs seems to offer a fix. But often these are only quickfixes, to lead to more problems later on. Skills like programming, database management, GUI development, project management, etc are not syntactic. Without strong conceptual foundations and clarity of concepts, you become only a coder, not a developer (in the case of programming). And university courses teaching such topics often do not provide the necessary depth in this area. The external diploma programs also often leave this to the University, and focuses on shallow exposure to the various language constructs. In the process, you may remember what a complex construct may denote, but are unable to write even a 10 line program, given a problem. I believe, in most languages, you need to really learn only 30-40% of the language for most practical requirements. The rest is best learned while working with the language (for project, job, etc). Listening to lectures enumerating library functions, or even attempting to remember them is not a productive way to spend your life!

A related aspect is the focus on tools. Exposure to and working with tools is important, to ensure you can produce solutions. But in a field like IT, tools change rapidly. Fortran was a popular general programming language, which gave way to C, then C++ and today Java. Many databases have come and gone, and many are still in the fray. Different companies, and different projects may use different ones. Development environments and frameworks are coming in almost daily. Therefore attachment to a tool is not healthy in the long run. The ability to move along the range of tools as they come along is more important. And this necessitates a good foundation in the respective area. Shallow exposure to a lot of buzzwords may get you a job of some sort in the short run. But you lose the opportunity to grow, given the limited adaptive ability.

Tools are afterall just tools. They facilitate work. They are not solutions themselves. Mastery over the intricate functions and options in a tool, is not worth investing your time on. Effective use of the tool to solve your problem should be the focus. So when you get lectures on which button to press for activating various functions, remember your time is getting wasted. The focus of teaching a tool — if at all, such a thing makes sense — should be to make you aware of the broad functionalities, and limitations of a tool, so that you know what can be done well and what cannot be done well using the tool. The important knowledge is about the associated process (usually effective software development), and how and where a tool fits into that process.

Many courses violate all these. You feel you are a master of Java or .NET at the end of the course, but you can’t write an efficient program to solve a simple problem. Watch out for such disasters.

I have seen many Java faculty who can’t explain basic concepts such as role of interfaces, how to use threads to solve a problem, need for inheritance, and so on. A faculty who talks nicely, need not be knowledgeable. Adequate access to lab is important, not enough to have a fancy lab which is open for you just 1 or 2 hours a day. Educational resources in the form of good library — digial and/or physical — are important. And above all, an ambience which encourages study.

As in all aspects of life, educated choices are important.

bike ka raja!

March 1, 2013

I am on a bike,
Driving as I like;
The roads are mine,
And I draw the line;

Signals are for you,
Not for bikers to view;
Use footpath for sure,
when it can’t us lure;

When me on my bike,
The world is as I like;
Your life, if you care,
On the road, not to dare;

If indeed you must come,
Whatever that may come;
Keep eyes all around,
And alert like a hound.


February 20, 2013

The end is now so close, feelings begin to grow,
What lies beyond? Hope it is not too sour.
What the escape is from, I do not know,
Was my path wrong? Or wrong image I wore?

An odd path indeed to take, frowns I can sure see;
Life needs a meaning, a target daily to seek;
The good of the world, isn’t all I should see?
Formulas of ego and greed, for me, are too weak,

[Feb 2013]

The little sasi all alone in the sky!

January 8, 2013


The little sasi, all alone in the sky!!

A picture taken while leaving from CDAC Kharghar office one day in the evening…

Innovation and Research….

October 11, 2011

These are days on Steve Jobs for the IT oriented, and Jagjit Singh for the music oriented — singing about them, thinking about them, blogging about them… I know little about music, so this is not about Jagjit Singh. I know a little about IT (after all I am a little sasi!), so that is what I am on to.

I bumped into this blog post — nice, not the usual advice useful in life. http://news.cnet.com/8301-13579_3-20117575-37/what-i-learned-from-steve-jobs/ Check it out.

Particularly interesting are the hear the experts, but dont listen to them. And the comment on committees. Far too often, in many forums, I have seen committees and experts making comments with little understanding and foresight. Collective opinion of committees with half baked experts, amount to not doing anything good. When one wants justification of cost, another wants market survey, and someone ‘knows’ this wont work. Intuition is what has led to most innovative things in the world history, not scientifically studied mathematically precise proposals.

But intuition is tricky to identify. It is easy to pass of something you have read somewhere, or done somewhere as your intuition. One way to judge your intuition is with your own passion for it, your belief in it. Someone else’s intuition can not determine your path. And secondly, intuition can not be arbitrary. A qualitative argument against usual objections should be possible. I remember someone making a proposal for project for voice biometrics. Finger print is understandable — there is enough belief in the world that no two prints are alike. Whether you can capture enough features to retain this uniqueness, and whether such features are adequately safe are different stories. At least finger print biometrics looks feasible. But voice is another story. Good mimicry artists can copy a voice even better than the original! That means, at least the human audible part of speech does not have enough uniqueness. So, there should be some conviction that the uniqueness is there in some aspect or other, before claiming to work on that. Such sanity checks balanced with some faith in intuition is needed for nurturing innovation.

I also liked the cue that build something ‘you’ will use. The intended target users may not be able to relate to the product, even when they see it. Somethings take time. But if it fills a need or utility that you have felt, then, perhaps, there is hope. It is the same thing that open source people talk of ‘scratching one’s itch’. If you visualise possible itches and build solutions, it may not sell. If you feel an itch, solve it, and that could be lessons (solutions) for others, and you would have developed it at least to solve your own problem.

However, all this is easier said than done. When you fund a project, how do you know what to fund, and where you think is money drain with no output? The usual solution is try committees to evaluate the proposals – through documents, presentations, etc. But finding people who can be honest to themselves, and respect intuition and be futuristic is not easy. Asking stupid and nasty questions is ok; but taking views is not. But, far too often, that is what happens. And not surprisingly, innovation takes a back seat….