I am hearing more and more about events. Starting from Twitter, to Facebook updates, to timeline makers, everybody realizes that events are important. And something needs to be done using events. I definitely feel that we are going to see and hear more and more about events and computational appraoches to deal with the complete eco-system dealing with events. Ultimeately, to model the world, one must use both events and objects. In computer science we have done well to deal with objects, but we really don;t know how to deal with events.

Multimedia has become a way of life in computing. Yet mulimedia computing is one of the most frquently misunderstood ‘elephant’ in computing where computer scientists look at it only from a very narrow perspective and consider that to be the real problem. Understanding how multimedia is interpreted and what are the implications of having mutimedia data in many aspects of computing is really essential. I feel that in many applications, but find it a bit frustrating that most people love to deal sith only a limired aspect of information in one medium and ignore other key sources of information from other medium.

Now it is impossible to bring in all these aspects in just one book, but at least one could try to make people sensitive to these. ANd if a simple step could be taken to make computer scientists, particularly incoming computing oriented people, then we will make good long term progress. This project is motivated by this desire. Don;t know how far we can go in that direction, but it will be nice to make as effort in this right direction.

NII has a very strong research program and is definitely one of the most exciting research places. Many research projects here are of world level. In my own research area, Shin’ichi Sato is doing very good work on video retrieval, concept detection in video, face recognition and related areas. In many other areas the work is exciting. I liked the research directions in what they call 3D internet and some aspects related to detecting emotions in text and endowing these emotions to avatars in virtual world.

It is clear, however, that research is quite conservative. Most research problems are the problems that become popular elsewhere before they are pursued by researchers here. I did not see any research project that suggested that somebody is adopting a new perspective or is addressing a new problem.
Similarly, it appears that there is some lip-service to working with industry and thinking of start-up but there are neither processes to encourage this not there is any environment to make this happen. Interestingly, most advisory board members seemed to be saying that this is somehow possible in USA but does not happen at other places so one should not worry much about this. I also found it surprising that in the society that believes in social harmony and social connections, in NII much emphasis is on making physical netwroks efficient but I did not hear any exciting research in use of Social Networks. The land of mobile phones, did not have any exciting research projects in this direction also.

It does appear that the research problems and research directions in USA are usually more aggressive and bold. But then, even in USA most universities are as conservative as NII researchers here and many academicians at strong universities will never think os doing ‘applied research’ till a new field of applied research creates many interesting ‘basic research’ directions. I guess, the situation is usually that of the drunk man under the lamp post — that I posted here recently.

The toilet seat has hot and cold water, variable rate bidet and shower. This was there the last time I was here. This time, it has pressure controlled exhaust fan and deodorizer. As soon as you sit on the seat, the exhaust fan starts and the deodorizer gets in action. So you don;t have to smell any odors even in the toilet — not even the one that you are responsible for. I am wowed!

And I am told that Akihabara is a different place now then when I saw it 4 years ago. Eager to see that.

At one time, recommendation engines were lot more focussed at only a few shopping sites. But as advertisement started becoming dominant, search and recommendation also started coming close. But that is only a short term trend. In fact recommendation is central to solving the information overload problem. As we all know, information overload is one the major problems created by advancement in information technology, particularly the Web.

I am sure soon we will see recommendation engines become very active in helping recommend each piece of information for a user and become a strong filter. That will mean having a better information about the user and her context and use that in filtering what information should go to which user. This is definitely an interesting challenge for people working on reducing information overlaod.