We should do what kind of research

Speaker: John strong

Finishing: sunling

Why are some books, etc., it was not the "hold"? new maintainers ' Guide, in the paper the most of research topics, how to choose? "in the Microsoft Research Asia and a lecture" on Friday, in speech and pattern recognition field work by 25 of hua qiang teacher trainees were concerned about the above problem, based on their own for many years engaged in research and education experience, has been very beneficial to share.

Let us listen to Hua teacher to interns have their say in his experiences and adventures.

How to send Paper?: a study of four types and three metrics

Question 1: the same is done academic, why some people made a lot of paper, some people did a story?

Hua qiang, in fact, the concept paper, I went from scratch, then from there to no changes.

When I was younger, I didn't write the concept paper is herself doing their favorite projects and research; later, Ph. when found many teachers have a paper promotion pressure, I also have this concept; then later discovered that only concern the paper is wrong, but rather should be concerned with the study itself, so that now there is no concept of the paper. Paper is just a by-product of the study, there are more important to do research.

Based on the objective of the study and practical, I put on is divided into four categories:

The first is "Boer-" the creation of quantum mechanics, purely Boer stems from his love, such research is curiosity-driven research, but this kind of measure of whether the invention is "new knowledge".

The second category is "Pasteur", Institut Pasteur, hope to prevent infectious diseases, through a number of experiments proved the theory of the cause lies in bacteria, his research has resulted in new knowledge, and thus opens up a new era of disease prevention, directly benefit mankind.

The third category is "Edison", his research and theory, however, pay more attention to usability, rather than invent new knowledge.

The last category is the "academic garbage," neither practical nor produce new knowledge, researchers may not be interested in.

Unfortunately, this kind of research are likely to be a lot of paper, as long as you have so-called "game rules", I would like to know! in this sense, the firm does not do this type of study even if made by the Green paper, is not enough, resulting in a stack of paper waste, garbage is academic.

Therefore, we are in the process of doing research, you need to ask yourself: what I want to be the type of research, for example, I would tell myself to be "Pasteur" research.

A study of type, select the topic, there are three key issues: passion (Passion), excellent (Excellence/Greatness) and effects (Impact).

Passion is a deep feeling of the "itch", if not to do it, my heart will always be ready for it. However, make people passionate about a lot of things possible, we need to think about their own ability is reached, your own in this area can do the best, we must learn how to measure your own location. The "impact" varies from person to person, for the Professor, the contribution to human society, science and technology progress of contribution size; for persons employed in enterprises of the researchers, in addition to the above considerations, the research results on product and business contributions can also be used as an internal impact indicators. All in all, there needs to be measured.

In considering the clear four types, three elements, and then, if you can continue to work hard, the paper is not a problem.

Imagine your research and useful, and the generation of new knowledge, and are the best, but very influential, but have not out?

Have the opportunity to do his enthusiasms like research is a very, very luxurious.

Issue 2: the logic of some paper is correct, but it is not based on a very practical basis of assumptions, we should think of this problem?

Hua qiang, I teach at the University of Hong Kong, was responsible for the Department of computer science undergraduate program design and management.

Science and engineering of difference and relationship, I at each session students enrolled must say lesson:

Excellent engineering company, first of all to be excellent scientists.

Some science behind closed doors, "may", in particular in the framework of the system, logical consistency, the scientific reasoning, can produce new knowledge. Engineering company you need to find solutions to actual problems, in logic, you need to consider whether the model assumptions and practical. For example, if the only bridge design-8 typhoon, Typhoon encountered 8.1, bridge collapse, which would issue layer out, so it must be in accordance with the actual situation considerations design. Science aims to understand the essence of things, but the project you need both to understand the actual situation, also invented something unprecedented. Thus, if you want to be an excellent engineering company, it is necessary to have a solidScientific knowledge, but also for practical issues of insight. Think about it, "School of the University" in five years, very lucky, undergraduate education is very solid, we must learn many basic course, encourage students early into the laboratory involved in the actual research project, is the so-called "lishi blend", is engaged in the engineering research has laid a solid foundation.

For novices, academic navigation--you and your research "

Question 3: I want to be really useful things, but our predecessors have many squabbling, find it difficult to start, what should you do?

Hua qiang, in fact, great people is to put the puzzle are put by our predecessors, this is the best time to make a contribution.

Issue 4: research directions recognized bottlenecks?

Hua qiang, first of all, recognized the bottleneck is not necessarily true bottlenecks, progress in the field of research is always a high tide is low, the patience to delve into.

Secondly, you want to know your area important question is what, and then see whether you are trying to solve those problems. If not, ask yourself why. As long as the general direction is right, left, is to the courage, effort.

Issue 5: research new maintainers ' Guide, you have any suggestions?

Hua qiang, most students into, I must first "brainwashing" (as in issue 1 answer), and help them think clearly want to do what kind of research.

And then give them enough patience and encouragement, until they are able to have independent thinking, independent research. I was reading Dr. period, under pressure, trouble don't be nervous, in fact, now on the podium in his schooling and work also met a lot of difficulties, the so-called "get a go".

In addition, we must learn how to maximize use of your environment.

For example, I'm studying and teaching at the University of Hong Kong, the library every Wednesday there will be new releases, those books on the shelves until the library in a corner of the lobby, so I am Wednesday morning as possible first to see the new book, in just one morning, I can generally browse all subjects was added to our catalog on the new book, the first time the useful new book borrowing from a good read. So stick to more than ten years, the harvest is abundant. Then I put the "know-how" tell my students, some students follow, as a result I occasionally cannot borrow the first time I'd like to borrow new book.

Question 6: in your requirements we will read "you and you (" You & YourResearch ")," this lecture in say, those first-class researchers, they always said he was young enough lucky.

First, because confident doing a first job, so only after work to get more exciting; secondly to work under pressure, to make a good job; Moreover, due to other people's expectations, we can make beautiful work. I would like to know your research process, have encountered some difficulties, is how to overcome? confidence where it came from?

Hua qiang, indeed, difficult very much, ten things, also a two piece is "decent", a lot of my own unsatisfactory results not published, so we don't know.

About self-confidence, I actually lucky, from small to large school period has been the first place, so there's a natural confidence. But, in fact, the second name, first name in the future maybe difference is not very big, critical or personal pursuit, success depends on how the ego-driven. Back to the above seminars, Turing Award laureates RichardHamming "you and your research" ("You & YourResearch"), I asked my students must read the documentation, here to share his on some of the great men of the common thinking, thinking how great they are. The IQ is circumstances, why some people are magnificent, but some have sabotaged? the logic is simple, but to personally carry out, but very difficult, I myself often revisit this lecture. Here, Hamming cited example, Hamming at Bell Labs, surprised than his years of light JohnTurkey so genius, so I rushed into the then heads of Bode's Office to ask: "like my age, how can the same knowledgeable JohnTurkey?", the results Bode said: "as he was for many years of hard work." The accumulation of knowledge corresponding to bank deposits, compound interest can accumulate more debt, more. Every day I asked myself the question: is there any progress today, if not awake. The result is serviceable if guarantee every day, ten years later knowledge not profound or difficult.

Remind students that if the research as a career, it is to be very careful, because only a few were able to put.

Of the study is that although very hard, but when you find and solve problems, the satisfaction of unspeakable, other are not important, this feeling itself suffice. Secondly, there must be a "pure heart", for example, ask yourself whether the "progress" of the day before going to sleep. At the same time also have "IQ + EQ + q" spirit in the face of failure and difficult to maintain, but also have an optimistic mood, let the mundane utility will not bother you.

-------------------------------------------------------------------------------------------------------------------------------------------

-

Introduction: Dr. Fok

Dr. Fok in August 2007, he joined Microsoft Research Asia fellow voice group.

Prior to that in the Hong Kong University Department of computer science for nearly ten years. In 1995-1997 in Japan Kyoto "International Institute for advanced communication technologies" (ATR) engaged in speech recognition research and participated in the world's first oral machine translation project. Over the past 20 years, he has always insisted on, for speech recognition, handwriting recognition, optical character recognition, based on the biometric user recognition, speech image processing and computer vision hardware design, make an important contribution to the field.

Welcome micro- concern soft Asian Institute micro-Bo: http://t.sina.com.cn/msra

Related reading:

Ignited the passion for research

Shaping the research for professional career

Study of the N-word Scriptures