Home » For Students

Research Advice

Submitted by Steven Chan on Sat, 08/27/2005 - 11:07am.

FORCE: Finding On-campus Research at Cal Easily

UPE Officers

officers@upe.cs.berkeley.edu

ABSTRACT

Doing research can benefit you now and your future. You earn money, enhance your knowledge, and build relationship with professors from whom you get the recommendation letters. Later on when you graduate, it helps you get into graduate school and makes your resume look good. This article is a survey of getting Computer Science (CS) research for undergraduate students at University of California, Berkeley.

1. INTRODUCTION

Students in all years are eligible for and capable of doing research. When you are a freshman, you try hard to adapt to the new environment. Not many students get their research in their first year. Don't make it as your excuse of not doing research, though. You are eligible for doing research regardless of what year you are! You will stay here for 4 years and have the greatest potential among all the others. If you are a sophomore, you are the ideal candidate. You will stay here for 3 years - a pretty long time, and you should have become used to the university's life and have sufficient background to understand and particulate in the research area in which you are interested. Juniors, please do not feel that you are too old for research. You have taken all lower-division classes or some upper-division classes. Compared to the 1st and 2nd year students, you will spend less time to bring positive contribution to your research group. Seniors should not think that you are too old, either. It is possible to become a full-time paid researcher for a year and apply to graduate school.

A lot of research opportunities for CS students are out there. In the CS department web site, the "Projects" hyperlink links you to a list of on-going research projects. Research programs, such as URO, URAP, SUPERB, are designed for you. A fact not many students know is that other universities have summer research programs and encourage Berkeley students to apply. If you don't like filling in application forms, talk to a professor directly. Scared of professors? Talk to your teaching assistant (TA), lab assistant, or other graduate students. They would love to help you. The officers of UPE, the computer science honor society, are always welcome to be your peer advicer. Another honor society HKN provides similar academic service as well. The various ways to doing research will be introduced followed by case studies intended to give you an example of how other people succeeded.

2. CS RESEARCH PROJECTS

The research projects we talk about here are those created by a group of professors and graduate students at UC Berkeley. The professors and research students set up an environment to verify their hypothesis and discover new theories. For example, if they would like to find a good strategy in playing blackjack, they might write a program in which there may be 10 players in the game. Each player would play in different strategies. Then the computer program would run and play 1 million games to find out which player wins the most. After that, they would do the stimulation again with 2 players, take data, and publish the results. There is no deadline for applying, and the research project is fixed.

Here is a list of some CS popular Research Projects. [1]

  • OceanStore
    • http://oceanstore.cs.berkeley.edu/jobs/index.html
    • Topics: Networking. Some examples include Detecting and Resisting DDoS via P2P, Peer to Peer (P2P) Network Composition, Deploying and Building Application on the Bayeux Multicast System, Service Discovery Service on Tapestry, etc.
    • Knowledge: Network protocols, Java, Algorithms, etc.
  • Group for User Interface Research (GUIR)
  • Optical
    • http://www.cs.berkeley.edu/Research/Projects/optical/
    • Topics: This project is studying the cornea, the front part of the eye that is responsible for three quarters of the light refraction in the eye. Visual correction through contact lenses or recently developed corneal surgeries require precise information about the shape of the cornea. Through the use of geometric modeling and scientific visualization, this project is developing sophisticated techniques to provide improved shape representations of patients' corneas.
    • Knowledge: CS 184 recommended.
  • Service Architecture for Heterogeneous Access, Resources, and Applications (SAHARA)
    • http://sahara.cs.berkeley.edu/
    • Topics: Our goal is to understand how to create end-to-end telecommunications services with desirable and predictable properties, such as performance and reliability, when provisioned from multiple and independent service providers. We are developing a new architecture for future telecommunications services that supports the dynamic confederation of sometimes collaborating and sometimes competing service providers.
    • Knowledge: Unix experience and basic networking and/or knowledge of network protocols. EE 122 recommended.
  • The Berkeley Initiative in Soft Computing (BISC)
    • http://www-bisc.cs.berkeley.edu/
    • Topics: Fuzzy logic and Artificial Intelligence (AI). An ongoing project is applying Fuzzy logic to a search engine to get better results.
    • Knowledge: Not much required. Fuzzy logic and/or CS 188 are recommended.

CASE STUDY: Allen, Senior, BISC Research Assistant. [2]

Allen has done research in BISC for 2 semesters. He heard BISC from his friend Harman last year, and attended the first meeting. Dr. Masoud Nikravesh was happy to have him in the group, and after a few meetings everybody in the group split up into separate project groups. Now his group is led by a GSI. His secret of getting the research is revealed from what he said, "If you talk with him (Dr. Nikravesh), he will be happy to let you do work." BISC requires only experience in C and Java from newcomers and expects new researchers to learn what they need to know. Interested students may attend classes held by Dr. Nikravesh every Thursday that discuss the concepts of fuzzy logic, neural networks, etc. Allen thought that important things a student should have are time and interest. He said, "Research usually takes more time than what the units indicate, and you'll be bored out of your mind if you don't like the subject, so be sure to pick a research project that you know you'll like and stick with it."

CASE STUDY: David, Junior, Undergraduate Student Researcher of the First Order Probabilistic Logic (FOPL) group. [3]

David did his AI research under Professor Stuart Russell in his sophomore year for 1 semester. He knew one of the graduate students in the FOPL group. After approaching the graduate student and expressing interest in joining, he spoke with Professor Russell and was able to join the group. The undergraduate artificial intelligence course, CS 188, is critical for getting his research. In addition to CS 188, he recommends interested people to take other probability and statistics courses, at a minimum Stat 25, preferrably a more advanced upper division course, before joining his group. David shares the same opinion as Allen about what attitude a student should have in doing research. He said, "A student should be enthusiastic about the research area and very willing to do independent reading in order to learn necessary background material."

3. RESEARCH PROGRAMS AT UC BERKELEY

Research programs differ from research projects in the way in which participants do research of a small project in some period of times, usually 8 weeks to 16 weeks, and they may not continue doing the projects afterwards. Another big difference is that like every other applicants, you need to submit various documents, such as your transcript. A research program also has a collection of research projects and a deadline for accepting applications. When applying, you can indicate which project you are interested in working on. The programs' web sites will tell you what documents are required for applying.

All CS Research Programs of UC Berkeley are shown next.

  • Undergraduate Research Apprentice Program (URAP)

4. RESEARCH PROGRAMS FROM OTHER PLACES

Research programs from other places that accept Berkeley students are almost always summer programs. You will stay in a new place for the whole summer and conduct research, make friends, study, and have fun. Summer research programs are offered by not only U.S. universities and companies but also foreign instituations. They are similar to SUPERB - it has deadline and lasts 8-12 weeks in the summer. However, you may not be eligible for every program. Applicants of some programs have to be a U.S. Citizen. Different programs have different standard of eligiblity of an applicant and requires different things along with the application forms, so check out their web sites for more details.

Below are some CS Summer Research Programs from other places.

  • Research Experiences for Undergraduates (REU) program
  • Summer Research Program at CERN (Geneva, Switzerland)
    • http://www.um-cern-reu.org/
    • Deadline: Early January of each year.
    • Eligibility: U.S. Citizens and permanent residents only. Students must be registered at an accredited 4 year institution and be in their junior or senior year.
  • Summer Research Program at University of Colorado, Boulder
    • http://www.cs.colorado.edu/reu/
    • Deadline: Early March of each year.
    • Eligibility: U.S. Citizens and permanent residents only. Applicants must be enrolled in a degree program.

5. DO RESEARCH WITH A PROFESSOR

Doing research with a professor directly lets you get to know the professor quickly. However, you should be pretty good at the research area with whom the professor you do research. Most people get their one-to-one research with the professor by first visit their office hours. Be sure to have something interesting to say, though. You may read some of the professor’s papers, ask if you don’t understand a part of it, and express your ideas and opinion on his/her papers.

The research topics along with the professors whose interest matches are listed below. [4]

  • AI
    • Professors: Russell, Wilensky.
    • Took 188 with grade of A or better, or you have done some research in AI.
  • Graphics
    • Professors: Barsky, O'Brien.
    • Took 184 with grade of A or better, or you have done some research in Graphics.
  • Database Management Systems
    • Professors: Franklin, Hellerstein.
    • Took 186 with grade of A or better, or you have done some research in Database.
  • Human-Computer Interaction
    • Professors: Canny, Dey, Hearst, Landay, Mankoff.
    • Took 160 with grade of A or better, or you have done some research in User Interface.
  • Operating Systems (OS)
    • Professors: Brewer, Joseph, Smith.
    • Took 162 with grade of A or better, or you have done some research in OS.

6. CONCLUSIONS

Almost all students who would like to do research are able to find research. We have found that successors start early to look for research, do not do very bad in the courses whose topics are similar to that of the research groups to which they apply, and are aggressive and talk to a lot of professors/graduate students. Undergraduate student researchers join research groups, apply to research programs, or do research with a professor. We wish you have success in finding your research.

REFERENCE

[1] Computer Science department of UC Berkeley, "CS Research Projects" [Online] 17 Feb. 2004. <http://www.cs.berkeley.edu/Research/Projects/>.
[2] Cheung, Allen. Personal Communication. 13 Feb. 2004.
[3] Sontag, David. Personal Communication. 16 Feb. 2004.
[4] Computer Science department of UC Berkeley, "CS Research Areas" [Online] 17 Feb. 2004. <http://www.eecs.berkeley.edu/Research/Areas/CS/>.

‹ Lecture SeriesupResearch Classifieds ›
» login or register to post comments

Warning: Unknown column 'cache' in 'field list' query: INSERT INTO sessions (sid, uid, cache, hostname, session, timestamp) VALUES ('58ec199219d911a7cf3e9b46143f6442', 0, 0, '69.237.252.182', '', 1170847439) in /home/raid/upeweb/public_html/includes/database.mysql.inc on line 120