At the Point from 2005 to 2006

I had hesitated that whether I should write somethings quite private (always speak to myself quietly or always write them in my locked dairy secretly) in this public place. One point I could not make is I don't want what I thought in my mind and the deep of my heart known by the others, in particular I contribute to let them known, which is what I am unwilling to do ever!

But, why did I eventually decide to type those into this blog? The reason is quite simple! I expect to make some changes, especially in this special season! I don't want to judge whether this decision is right or not while sometime I still ask myself about this decision. After I made to do it, I will never regret doing it and will insist on it. I believe that this will keep trace of what I have thought, what I have achieved, what I have dreamed, what I have planned, and so on. It will become a record to present every piece of fractual images of my life to myself!

What I write here are not only what I have achieved but also what I have failed to do. The most of all, I anticipated, is what is the experiences and lessons learnt from those unchanged facts. They are precious to my future path!

2005 is my first year of Ph.D. It is very important with respect to the fact that during this year I should lay well out a series of personalised research methods, and find an enhanced and fundamental standing-foot in the interested research area which is crucial not only to the Ph.D study but also to the further research career though it maybe has some minor changes later. This was my main job I defined and assigned to myself when I just started Ph.D research. In the following I will summarise this one-year time based on two categories: Research and Life. Here, I will not say too much about what I was successful to make because that is always the first thing popping up to my minds. What I am focusing here is what I should have done and the lessons learnt from those.

1. Research

According to the main job I have described above, basically, I am satisfied with the whole process. It could be reflected by the result of nine-month report. Both supervisors are very happy with my process and the results currently achieved. However, within my original plan, what I should do was more than that. I came to draw some reasons as follows:

a) Time management: in the beginning, in order to find a research area best fitted, I spent too much time in reading the relevant literatures. More seriously, the range I depicted to find the target was too large. I included too much information at one time which made me lost during the reading. The reading is very necessary and it is a routine work, but that is not the whole of the research. The role of the literatures is to provide useful, helpful, and supportive facts to backup your ideas, also to give you magical thoughts from the large amounts of relevant information clusters. Eventually you have to draw a personalised conclusion for your research work.
b) Multiple sessions: I think this problem is my old habit. I had thought for a long time to change it, but always failed. I was used to act step by step as original plan without any interruptions. That was good. But sometime you have to allow plan to change due to certain unexpected reasons. The lack of the ability to deal with this stuff flexibly was my problem. I have made a great effort to get over, though the obvious progress has been achieved, that is not enough. I need to enhance this ability to cope with multiple tasks simultaneously.
c) Action is first: I am very proud that I have a strong ability to plan and organise well. During this year, I found out that after nice plan, I need to act swiftly and effectively. This is very crucial for the research, I think. For example, the works you are performing now maybe is under the research by the other group of researchers, who is the first one to publish and represent the works, he/she will have absolute power in this field if the works are agreed upon. Time is crucial issue at this point. Thus, action swiftly and effectively on well-designed plan is important. In addition, new findings will be unveiled after a series of actions, for example a set of experiments. The new findings are more convictive with the support of the experience. So, don't imagine to get the new findings only by reading a large amount of papers.
d) Get the fresh ideas from the colleagues: I understood that don't save your questions for yourself. Share them with your colleagues, you will get a surpirse, not only the answers about your questions themselves, but also the correlated and relevant problems. You will learn more than what you expected from the questions.
e) Keep regular contact with my supervisors: this is not my problem. I believe that I did well in this aspect. Here, I just want to warn myself that you should get more and more knowledge out of from the minds of the supervisors. That is a precious opportunity for you to learn easily.

Till now, the problems identified above were just what I could figure out. It must be some potential problems I could not take them out. This list is not exhauted! With the time being, I will be very happy and exited to see it becomes shorter and shorter.

2. Life

I don't want to speak too much here about my life. Only one thing, I hate it very much. Unregular life style! I always want more time to learn the knowledge. Besides the study time in the university, I will spend loads of time at home on the casual reading task which is not quite relevant to my research, but increase my knowledge-base, in particular at night. As a result, it directly took up my rest time, then brought the effects to the next day's work. That is too bad! I know.

I must get rid of it, back to normal life! Take every seconds in the university putting them on the research. Arrange part of time from the work reasonably for the casual reading task. Make sure healthy sleeping time and regular exercise schedule! (Anyone happens to find out that I do not obey what I said above. You will be rewarded a feast in the luxury hotel. You define! )

The Scenarios for the Semantic Web

The Semantic Web has been a very popular topic in the popluar press and technical and academic resesarch literatures. This promising extension to the current Web has raised a number of expectations. These expectations could be potrayed as three different perspectives of the Semantic Web as follows:[1]

• a universal library, to be readily accessed and used by humans in a variety of information use contexts;
• the backdrop for the work of computational agents completing sophisticated activities on behalf of their human counterparts;
• a method for federating particular knowledge bases and databases to perform anticipated tasks for humans and their agents.

[1] Marshall, C. C. and Shipman, F. M. 2003. Which semantic web?. In Proceedings of the Fourteenth ACM Conference on Hypertext and Hypermedia (Nottingham, UK, August 26 - 30, 2003). HYPERTEXT '03. ACM Press, New York, NY, 57-66.

Bottom Down and Bottom Up

Categories systems undoubtely make the data, information and knowledge easily to combine and communicate each other. They are two kinds of categories systems:

• Bottom down categories: the normal categories like Google's and Yahoo's.
• Bottom up categories: emerging tagging systems, user-generated labels created in free association with an object like a photo or a webpage.

Semantic Web Comes

The widespread use of internet technologies has created a wave of innovations and interactions leading to a major impact in almost every areas of human activities, albeit a large number of these applications were not foreseen thirty years ago, when the Internet architecture was developed. Besides, in only 14 years the World Wide Web became the greatest network accessible repository of human knowledge. It contains around 3 billion documents, which are accessed by over 500 million world-wide users. Moreover, the Internet is continuing to evolve to integrate other technologies such as multimedia and mobile technologies. On the other hand, information overload and poor aggregation of contents make the current Web inadequate for automatic transfers of information. As a consequence, the current Web is evolving for a new generation called Semantic Web, in which data and services are understandable and usable not only by humans but also by computers. "Moreover, the Semantic Web may further evolve to a Sentient Web, which is a further new generation of Web with capabilities for sentience. As a consequence, new technologies and concepts are being developed and new ways of utilizing or integrating older and new technologies are being constantly developed." [1]


[1] Mário Freire and Manuela Pereira, Eds., The introduction in: "Encyclopedia of Internet Technologies and Applications", Department of Informatics, University of Beira Interior, Portugal

Challenge to Traditional Software Engineering

Microsoft and Google have combined with Sun to fund an new Lab in University of California, Berkeley to the development of new approach for software development.

In traditional software engieerning, work is completed in sequential stages starting from system concept design to development, assessment or test, deployment and operation. This approach is often too slow. Instead of infrequent, well-tested upgrades, code for internet services is continually being modified on the fly. This fix-it-as-you-go approach enables speedier deployment, but it also requires a large technical support group to make sure operations are not disrupted as bugs are resolved.

The message from Prof. David Patterson in UBC, the founding director of this lab: "right now, it takes a large company employing hundreds of really smart people to support internet service". The new approach from the lab is to develop technology that eliminates the need for such a large organization, opening up innovation opportunities for small groups or even individual entrepreneurs. It is obvious that the big software companies, such as Microsoft and Google, welcome the birth of this new approach. Thus they could save lots of human resources and financial resources for the company.

According to Prof.David Patternson, the new approach could achieve this goal by applying statistical mahine learning - the same technology used successfully in the recent antonomous vehicle grand challenge - to the development of computer systems.

Disciplines in Artificial Intelligence

I have seen a discussion about the disciplines (a.k.a. the academic fields) that comsist of the strong AI research (strong AI is the implemenation of real intelligence; weak AI is the mimic of the intelligent behaviours) on the usernet comp.ai.philosophy. Summarising the discussion, there are seven distinct disciplines as follows:

• Philosophy
• Mathematics
• Computational Neuroscience
• Electronic Engineering (a.k.a Eletronics and Computer Science)
• Information Retrieval
• Cognitive Psychology
• Computer Science (emphasis on Artificial Intelligence)

To my knowledge, what I am wondering is whether the researches related to robotics are missing here. If not, which one or more of the disciplines above spots it? Or should we list it as the eighth? Robotics is the first impression for the people to understand the disciplines/technologies within the AI research. Although the research of robotics include many of the disciplines above, we could not decouple every piece of them. If that is case, I think the list should have only one item, i.e. mathematics, because mathematics is foundation of all of the disciplines listed, even for the robotics. The our discussion about the discipline in the Strong AI research is meaningless.

Pizza Order in the Semanic Web?!

Untamed fever has been annoying me for nearly 3 weeks. That is too bad! I even could not do any work at all. At this time, I really understand the truth value of health, in particular for a young energic man. I have no any ideas how the virus could visit me and stay here for so long time this time. For the sake of God, it happens during Xmas, this should-be-lovely season:(

I have to choose reading as my normal life because I really feel that is only thing I can do except eating the medicines. Occasionally I picked up this piece from some website (playing with sound please). It gives me a really strong injection at this time. One thing, it is very happy to see this kind of application people could imagine in our future life, which combines many technologies together to deliver the services in such an intelligent fashion. The customers would not only get satisfactory services but also get considerate "unexpected" care. That is so cool! But on the other hand, I doubt how many people are willing to accept this application. It seems to integrate too many sensitive data into the applications. In the other words, too many sensitive data should not be open in such "wildness". The Semantic Web make the existing web to be a hugh data storing base. The applications built upon it could communicate and understand each other with the great assistance of the large amount of available structured data. However, it does not mean that everyone could not easily hold the control of the data without the authorisation, in particualr the sensitive data. Taking this example, I believe that nobody expect that their medical records are unveiled for the public, even used by the third party in their real applications. Sound horrible! Maybe I exaggerate this issue.

US Computer Science Eager for Funding

A report issued by the Defense Science Boards earlier this year describes computer science as having outgrown the Defense Department's capacity to support and fund the industry it largely created. In the absence of a transition strategy, the possibility of the United States losing its competitive edge in university research is now very real. Over the past four years, DARPA estimates that it has cut funding for university research for computer science by almost half. In congressional testimony, DARPA director Anthony Tether said that some research projects have moved out of universities and into industry, and described DARPA funding as having remained "more or less constant." At the NSF, funding has actually increased, though the portion of the proposals for computer science projects it sponsors has dropped from roughly one-third to 14 percent. "We are looking at a situation where perhaps 40 percent of the good proposals we get, we don't have the money to fund," said the NSF's Peter Freeman, citing the war in Iraq and natural disasters as higher priorities for government funding. The lack of university funding has shifted much of the burden for supporting research to the corporate realm, which typically only supports short-term projects that have an evident business value. Since 1999, MIT's computer science department has seen the portion of its funding supplied by DARPA drop from 62 percent to 24 percent. As research funding becomes a lower priority for a government grappling with an escalating budget crisis, there is widespread concern that other nations, particularly China, could supplant the United States in the next 10 years as the world's leader in technological innovation. Evidence of this trend can be found in the facilities that many companies are establishing in China, India, and other countries, as well as the appearance of large new universities, such as one China recently opened with a capacity for 30,000 students.

[Source: ACM TECHNews Volume 7 Issue 873, December 2, 2005 ]

New Home is Coming!

Here will become my new home for my personal website. The original one is heavily out-of-date. I had thought for a long time to re-build it. What's more, I will blog and update my flying thoughts or any ideas from the research work at any time. Very nice! The only thing I am worrying about is the site is not my product. I could not control it totally even if I could edit the template. I don't wish that there is something wrong with it some day. Is that the really giant shoulder I could lean against?

Recently there have been a lot of discussions related to the fact that whether the blog could replace the traditional personal website in the era of Web 2.0. I have no clear ideas about it. At least I could not see any down sides of the traditional personal websites for Web 2.0. What I like of the blog is you could write the news(information) in the format of chronological records. That is one of the fashion I think helping to track the thinking process, which is very useful and precious for the academic research.

I have three blogs combined together to form my new personal website. They are David Liang | Home on the Internet, David Liang's Research, and David Liang's Knowledge Repository. More sites could be added later. Currently I will maintain these three. The descriptions of each site could be found in the anouncement section of the sites. The migration is still in the process!

Turst through Internet

What about trust between strangers who have no contact except through cyberspace? EBay claims more than 100 million registered users who buy and sell millions of items each day on the Internet.

Researchers at California Institute of Technology in Pasadena, Calif., and Baylor College of Medicine in Houston used hyperscan functional magnetic resonance imaging to study the brains of subjects 1,500 miles apart engaged in a trust-based financial transaction game.

“Hyperscanning is the simultaneous scanning of the socially interacting brains of two or more people,” said P. Read Montague, a Baylor neuroscience professor. “We were very interested in how trust evolves when subjects are not near one another and anonymous to each other,” he said.

Montague and colleagues Brooks King-Casas and Steven Quartz measured how and when their subjects made trust decisions by measuring changes in blood flow where the “intention-to-trust mechanism” occurs in the brain.

Over time, subjects became less reactive and more proactive in their willingness to trust one another.

“The most intriguing results were that the intention to trust shifted in time and developed in a part of the brain that also does very basic reward learning,” Quartz said. “We have shown that trust, one of the highest-level concepts we possess, is a reflection of basic brain mechanisms.”

There are many different aspects of trust research activities. For example, happy people are more trusting than the sad people; altruism increases the trust; trust behaviours are correlated to naturally increased level of reproductive hormone oxytocin in the blood (neurobiology aspect of trust), etc.

[Trust Game--players exchange certain amounts of money they are initially given with the possibility of a greater return from their partners. Researchers study how much a person is willing to risk in assuming the other will play and pay by the rules — in effect how much trust can be created between the players.]