Assignment 9

1) please interview your advisor (for approx 1/2 hour) what she/he thinks about being a PhD student in CS at CU and which expectations she/he has

2) post the two most important arguments which surfaced during the interview on this site!

Advice: schedule your appointment as soon as possible!

(you should not expect that your advisor will be available on a very short notice)!

Dan Knights - Advisor: Mike Mozer. I enjoyed having the opportunity to chat with Mike about these topics, and I feel that this was a useful assignment. Mike's two most important pieces of advice for a Ph.D student were as follows. 1) Do lots of research. Mike feels that it is important to do research projects in at least two different areas. This shows a future employer that you are at least somewhat multi-faceted, in contrast to someone who focused on only one topic throughout their entire program. These projects should lead to publications if possible. Either way one should leave one's program with several publications to one's name, especially if one is interested in a career in academia. And, 2) do lots of reading. According to Mike, many students retain from their college days a tendency to over-emphasize grades and classwork as the main measure of their success, and they should spend more time reading papers and focusing on their research.

Guy Cobb - Advisor: Xiao-Chuan Cai. 1) One of the more important aspects of being a Ph.D. student, especially at the beginning, is to have a good understanding of the research being done in different departments that might be of interest to you. Because of the strong interdisciplinary nature of much research, a potential area of interest might come from another area such as mechanical engineering, rather than originating in the computer science department. A good way to investigate something such as this is to work on a project with that group in a semi-official capacity to see what the research involves and to help gauge whether this are is something that might someday contain a potential thesis topic. 2) The semi-official nature of working on such projects means that a greater personal investment is involved than would otherwise be the case. Most likely this will be done in addition to the coursework and TA duties for the first semester or so. A fair number of late nights will be required for something like this, but that it will most likely be worth it if initially there seems to be a good chance that a thesis topic can be isolated in the course of the work.

Yifei Jiang – Advisor: Xiao-Chuan Cai. The most important topics I talked with Professor Cai are PhD research and teaching skill. 1) Professor Cai said the difference between the PhD students and master/undergraduate students is that master/undergraduate students are required the development ability but PhD students need strong ability of research, the ability of producing new knowledge which makes practical applications possible. He also mentioned the importance of publishing papers during the PhD research, specially the high quality papers. 2) Professor Cai said teaching skill is also an important aspect for the PhD students who want to work in university in future. PhD student should learn teaching skills from his teachers in the class and take TA position to get teaching experience.

Mohammad Al-Mutawa – Advisor Shivakant Mishra. Professor Mishra believes that the 3 years of doing PhD research are the most satisfying, in which a student can completely focuses on his or her research without worrying about other thing such as securing funds which is what many faculties and professors are distracted by. He said that one of the very nice things about being a PhD student in CU is that most areas of computer science are covered unlike many other schools which have few areas only. This coverage and diversity can be very beneficial even if the student is interested in a certain area. Also CU is a top class research university and it is visited by top class researchers and speakers. Professor Mishra expects the students to do cutting edge research that will have an impact. He also expects the students to be independent researchers that explore the area without being told what to do. Another thing he expects from the students is to start publishing at least a paper a year once in the second or third year of the program.

Note: Paul, Caleb, and Dola decided to interview Prof. Dirk Grunwald together (so as not to monopolize his limited free time). We initiated conversation with the questions included in the assignment.

Paul Marshall - Advisor: Dirk Grunwald. Dirk said he tries to focus on research projects that students are motivated by - and similarly, he tries to get students interested in projects that he has funding for. He said that he has found that the most success in grad school occurs when students have a real passion and motivation for what they are working on. He tries to put forth the same amount of effort that students are putting in, to identify problems and locate funding and opportunities, but it is up the the student to take advantage of those opportunities and take them as far as they are able.

Dola Saha - Advisor: Dirk Grunwald. As memtioned by Paul, I have seen him working with us (students) till midnight to submit a paper. He mentioned that Ph.D. program requires lots and lots of time and effort, and he expects his students to put in more and more effort in research. According to him, he has found that the most success occurs when students have a passion for the work they do, that's more important than what degree of funding or prior history in the topic he has. He also mentioned that if people can't find a passion for the material in their heart, then they are unlikely to be willing to put in the time needed to do an excellent job. That is why it is important to motivate students in research areas in which he has funding and also focus on research in which students themselves are motivated.

Caleb Phillips - Advisor: Dirk Grunwald. Dola and Paul did a pretty good job summarizing Dirk's answers. One thing they missed is that he said that he expects his students to work together on projects. This is actually one of reasons I favored CU among school choices - the systems lab in general has a great sense of collaboration.

Note: Lee and Jinho chose to interview Dr. Martha Palmer together to mitigate the impact to her very busy schedule. However, we still ended up talking for nearly two hours.

Lee Becker - Advisor: Martha Palmer. According to Martha you should only endeavor to get a PhD if there is nothing else in the world you'd rather be working on as the level of commitment, self-discipline, and self-motivation required to finish are too great otherwise. Specific advice she gave for success in Natural Language Processing is to build as much background knowledge in linguistics, and artificial intelligence. Even though this knowledge will not always be directly relevant it will give you a stronger understanding of the human brain which will better enable you in building NLP systems. More generally speaking, Martha stressed the importance of keeping fresh perspectives and maintaining a broad network of colleagues to keep your point of view fresh.

Jinho Choi - Advisor: Martha Palmer
1) Martha thinks the most important aspect of being a PhD student is self-motivation: students should be in PhD program not because it is something good to do, but because it is the most desired thing to do. An advisor can give guidance or make suggestions, but it is a student who is ought to be searching for their own research. In other words, students should have their own ideas and bring passion about them.

Martha also mentioned that she may get frustrated if students don't really pay attention to what they do. When an advisor leads students in some way, it is expected for students to make some progress. They should understand their tasks and make some kind of achievement in time.

2-1) I asked her that as a student who doesn't have much Linguistics background yet studying Natural Language Processing (NLP), how much Linguistics I should study during the PhD. She answered that it is not necessary for me to take all Linguistics courses, but at least the core ones such as syntax, semantics, and pragmatics. It is clear that the more understanding I have in Linguistics, the better application I can build for.

2-2) Another question I asked is that how the trend for NLP has been and how it will be. She said that NLP used be highly focused on syntax because people thought that syntax was the most practical way to approach the issues, but as time goes on, semantics are taking more and more roles. Once we have some solid research on semantics, we will move to work on pragmatics, that is our ultimate goal.

Overall, the interview is very helpful for me. I get to know Martha better and see how she could become so successful in this area. She shared her experiences for both uphill and downfall in her life, which gave me some kind of guidance in my academic life. Thank to Dr. Palmer to enlighten us in many ways.

Note: Michael, Shumin and Joel "interviewed" Professor Grudic.

Michael Otte - Advisor: Greg Grudic. Professor Grudic says that being a CS student at CU is an excellent choice, particularity if the student's chosen research area has a presence at CU. He adds that, under these circumstances, there is no limit to what the student can accomplish. Dr. Grudic expects his PhD students to teach him about the areas of their research that he is unfamiliar with, and also to point out the areas that they find surprising.

Shumin Wu - not much to add here.

Joel Pfeiffer - Same

Yuli Liang - Advisor: Dr. Jane Mulligan
Dr. Mulligan think, according to individual interest, PhD students could start from some small projects to have some flavor of research, learn the technical skill, and interact with people and academic community. During this process, students would become more confident and know what kind of research they want to do. At the end, besides the technique skill, she expects the student to have ability to communicate with other researchers and explain their research clearly to the people from different background.

John Michalakes - Advisor: Manish Vachharajani

Please refer to this link to access interview. Briefly, the two most important arguments were (1) utilitarian versus "other" reasons to earn a PhD, and (2) the role persuasion in research.

Note: Jeff and Jane decided to interview Prof. Katie Siek together to minimize the impact on her busy schedule.

Jeffrey LaMarche - Advisor: Katie Siek
Katie stresses that it is important to continually evaluate your plan of graduate work and to restructure things as appropriate. If the area you are performing research in is not interesting for you, then figure out why and attempt to find something that interests you more. If you do not love what you are researching, the work you will do will never be as good as it could be and could even lead to failure. It is also important to continually decide why you are pursuing a PhD degree. Katie also emphasizes balancing research, life, coursework, and everything else that is thrown at you while you are a graduate student. Although not always a perfect balance, it is important not to neglect your social life and other important aspects of your life. One thing that many people forget is to make sure to exercise regularly to maintain your body(which also has been shown to increase the abilities of your mind and decrease depression).

Jane Meyers - Katie also emphasized the importance of networking, whether this is by meeting people at colloquium sessions, conferences, local workshops, fellow students and faculty members or other students/professors at outside universities that are doing work of interest to you. She also mentioned how helpful it was to read papers of visiting speakers/professors before meeting them so that you can take full advantage of discussing similarities/parallels between your work. Because being a graduate student is more or less of a “lifestyle” it is important, especially in the first two years, to prioritize and multi-task well. Her suggestions include when appropriate “double dipping” (combining research and class projects), setting aside a fixed amount of time for email, setting weekly and semester goals, and having working meals/walks where you discuss projects/ideas with peers over food or exercise activities.

Soumya Ghosh - Advisor: Jane Mulligan
Jane is of the opinion that it is important for PhD students to get their feet wet in research early in their PhD careers. This according to her, comes from getting involved and customizing parts of ongoing research projects in the lab joined by the student. She also emphasized the importance of being able to communicate one's research to the community at large. According to her, it is important to talk to people from outside of one's area of interest, as you never know where the next big idea will come from. Finally, Jane emphasized that a student should approach his/her advisor not only when things are going well but more importantly when the going is tough.

Keith Maull - Advisor : Tammy Sumner
Overall this interview went well - I had about 10 questions lined up and we got through most of them if not directly. Tammy, incidentally was a student at CU as well - she received her PhD in '95 and Gerhard was her advisor. This was the source of an interesting conversation about what was exciting at CU then vs now. Some of the biggest take aways from the conversion were : 1) a PhD is merely preparation for the next stage in your career, so find ways that enhance that preparation and be tactical, 2) getting your money's worth is important, since doing this a tremendous commitment in time, energy and income loss, 3) for my contribution to her group I need to carve out a research area that is personally and professionally fulfilling while at the same time advancing the strategic goals of her research, 4) it is important to have honed skills in communicating the value of my work to others, particularly interdisciplinary audiences, and 5) picking 2-3 things to focus your work on is important to find interdisciplinary breadth, and that I should consider taking a quantitative/qualitative research methods course as well as some NLP/machine learning classes.

Saroch Panichsakul – Advisor: Richard Han
He expects his students to produce at least two papers to be submitted to high-quality and major conferences in their areas of research each year. This should be achieved at least in the last three years of their PhD study period. After graduation, he expects his students to do great things, be an independent thinker and contribute to both the research community and educational institution. Prof. Han said that CU is quite a good place to study. There are lots of opportunities here that can create creativity and there are lots of support from great fellow PhD students and faculty. He defines a successful PhD student as the one that can find his or her own topic, run with it, and succeed with it. He also mentioned Ahmol, one of his graduated students, that he is the one we should follow his steps. He believes that PhD study should be self-driven; no one can graduate without self motivation.

Kyu Han Koh – Advisor : Clayton Lewis

According to Dr. Lewis, a good school means not a school with reputation of excellence but the school where you can meet people whom you can learn from. If there were good examples or people you can learn from, it would be the best place for you. From this point of view, CU is the best school for me because there are professors whom I have wanted to work with and be advised by since I entered the graduate school and peers I can learn from.

Also, he said that being a Ph. D student means becoming an independent researcher. To be a successful Ph. D, Ph. D students have to devote themselves to research. Research is the fundamental and essential thing for Ph. D. We learn to do research by doing research. As a consequence of this process, you will become an independent scholar who can contribute the community as a functional member.

Holger Dick – Advisor: Gerhard Fischer

I had a rather free discussion about the PhD program with Gerhard and will try to reproduce two ideas that came up during this discussion and haven't been mentioned by other students. These are my own words and Gerhard cannot held responsible for misunderstandings.

1) Students should not only rely on their advisors and classes to get new ideas and/or feedback. Often times, collaborating with other students (e.g. creating a reading group) is of more help than a professor's opinion.

2) A PhD thesis should include something new, i.e., something that no one has done before. Thus, PhD students shouldn't rely too much on their advisor's tips but make up their mind on their own. They should try to move into new intellectual spheres; most probably these will be the same as the advisor's but that doesn't mean that the approach to the topics and the knowledge about theme should just be a subset of the advisor's.

Ashok Basawapatna- Advisor Jane Mulligan

Two notable things I took away from my discussion with Prof. Mulligan are as follows.

One is that students need to get their feet wet in both research and in reading technical papers as soon as possible and should have the ability to talk at length about their research. This allows the student to become exposed to various research areas.

The second is that a Ph.D. thesis may be the culmination of 2 or 3 smaller projects as opposed to one "groundbreaking" project. Thus, the more little projects a Ph.D. student starts to pursue the better prepared for a thesis the student is rather than searching for one enormous Ph.D. thesis idea.

Nwanua Elumeze - Mike Eisenberg.
1). The CS PhD should not be seen as an end on to itself, but as preparation for an exciting chance to participate in cutting edge research and educational opportunities in a wide range of fields, not just in CS. For instance, I had considered the degree programs in MAS or EDUC, but the CS is one of the most adaptable "union cards" for an academic position, and the range of research (even within the department) is quite wide.

2). Do little bits of writing, reading, building, experimentation, throw-away projects (constantly, if in little chunks)... you never know what you might learn or discover. It's also a good way to stay motivated without burning out. Be careful about collaborative efforts with other researchers, particularly from different fields. It is hard to do, and expectations might be different, so if you still want to, spell out all the terms from the beginning.

David Gnabasik - Advisor Debra Goldberg
Dr. Goldberg emphasized the motivation and persistence necessary to complete a Ph.D. At this stage, doing good research work is relatively more important than classwork. She has instituted weekly doctoral student group meetings in which everyone brings in a new paper and updates the group on their work. Along with attending this weekly meeting, she expects Ph.D. candidates to attend (if not present at) at least 3 conferences per year plus publish 1 or 2 (collaborative) papers per year. This includes researching and writing funding proposals.

We did not disagree about much, and talked most of the time about planning ahead sufficiently far in advance so that the right opportunities can be realized. The Ph.D. candidate has to be forthright about his/her present status in terms of work, attitude and commitment. In return, the candidate should strive to "rent space in someone else's brain" as a collaborative experience.

Rhonda Hoenigman - Advisor Liz Bradley
Liz believes that a PhD program is an apprenticeship for learning how to do research and how to pick research topics. The most important thing to get out of a PhD program in any subject is learning how to identify research topics, and distinguish good research topics from those that will lead to box canyons. Liz also believes that a computer science PhD program provides a way of thinking about a problem. While a person’s research area may not directly relate to topics such as compilers or operating systems, studying these areas reinforces the concepts of abstraction and representation, concepts that are critical to solving any technical problem. Computer science is a unique discipline in that it can be used, and frequently is used, to solve interdisciplinary problems, problems that lie on the boundary between computer science and another discipline. It is these types of problems that are fertile ground for intellectual discovery.

Last modified 30 October 2007 at 8:17 pm by RhondaHoenigman