c. Comment on the balance of teaching and research specialties in relation to programmatic need. How does the mix of teaching specialties reflect current and anticipated trends in the discipline, and demand by graduate and undergraduate students?
The research interests of the faculty cover most areas of Computer Science, especially those that cover the current most popular trends in the field. Our recent hire of Ben Hardekopf has ensured that we have adequate coverage of programming languages and compilers, especially in terms of our undergraduate offerings. Xifeng Yan has started a graduate course in data mining, an area of increasing interest especially among graduate students. The retirement of Terry Smith, does leave a gap, from an educational point of view, to cover Machine Learning, which is also an area of increasing interest among students. A hire in this area, as described in our development plan would complement our nascent strength in Data Mining as well as the research and educational demands in different sub-areas.
We have recently revamped our lower division undergrad courses. They have been redesigned and updated to reflect current industry needs while still giving students a broad skillset that will outlast any current trends. Another skill that needed by our undergraduates, especially when they pursue an industrial career is the ability to groups and to have good software engineering expertise. These skills are developed using pair programming in our lower division classes and several project-based courses, capped with our newly restructured capstone courses.
To ensure that our graduate students are exposed to the latest innovations in Computer Science, faculty are encourage to teach graduate 290's on their research topic, which give graduate students timely content. Our seminar courses,595's, are another way of discussing ad hoc and often inter-disciplinary the most recent research results. We also invite local industry leaders to teach courses that cover the latest trends and results, especially those with strong interest in industry, e.g, Thorsten Von Eicken, the founder of the successful cloud computing company, RightScale, explored and developed many of the foundational themes in Cloud Computing in a graduate CS 290 course as early as 2007.
d. What is the faculty instructional workload policy?
Full time faculty teach 3 courses each, usually two undergraduate and one graduate. Faculty are also encouraged, and usually do offer CS595 2 credit hour seminar classes.
e. Are faculty members sufficiently active in the pursuit of extramural funds in light of funding possibilities?
Yes, as the overall funding numbers indicate, CS faculty are very active, and pursue diverse funding opportunities.
g. Describe and evaluate the current system of governance in the department. Is there broad faculty participation in departmental governance and on campus committees?
The department is very democratic by nature and most critical decisions are made by general votes by the entire faculty. There are committees for undergraduate curriculum and affairs, graduate admission and affairs, recruitment (when there are hiring positions), public relations, diversity, acceleration and strategic planning. The department meets as a whole for 2 hour faculty meeting, typically averaging 5 per 10 week quarter, depending on the need. Critical recommendations made by the committees are presented, discussed and voted on by the faculty. This structure has worked well for our needs and kept the decision making process cohesive and collective. Due to the increasing size of the department, this collective decision making model is under discussion in the area of merits and promotions. Currently, all cases are discussed at a faculty meeting. The acceleration committee is a consultative committee, which is convened to give advice to a faculty requesting acceleration or to the chair. A year ago, this model was modified to require all merit cases be discussed during a single day, so called Merit Day, to provide more uniformity.
Faculty are active in the department affairs, and faculty meeting are well attended. On a campus level, computer science faculty have been active, including in recent year, leadership of the Committee on Academic Personnel and The Graduate Council, as well as membership in Planning and Budget, International Education, Research and Instructional Resources, the Graduate and Undergraduate Councils, and the executive committees of the College of Engineering and of Creative Studies. Our faculty has also been active in administrative roles, including Almeroth as Associate Dean of COE, Turk as Chair of Media Arts and technology and Chong as Director of Computer Engineering.
GRADUATE PROGRAM Self Assessment a. What are the goals of the graduate program(s), and what is the rationale for its content, structure (or lack thereof), and specialties? Discuss any graduate curricular problems and efforts to solve them. The size and composition of our graduate program has changed significantly in the last ten years. Our graduate program is about 108 Ph.D. students and 40 MS students, compared with 70 Ph.D. and 52 MS students in the 2000-01 academic year. We have intentionally grown our Ph.D. program while shrinking our masters program to create a more research-centric environment for our students, as well as to improve the quality of our program and its graduates.
The purpose of our Ph.D. program is to prepare students for research and teaching positions in universities and colleges, and for research and leadership positions in industry and government. The goal of the program is to train the students in the methods of scientific inquiry and independent research. Graduating students should be well instructed in the theory and practice of computer science principles. They should be able to plan and execute a research project, organize, analyze, and write the results, and present them clearly to a knowledgeable audience. This is accomplished through advanced course work and active participation with the faculty in their research programs. Our MS program is mostly a terminal degree program, with the goal of training the next generation of information technology professionals. However, in some cases exceptional MS students transition into our Ph.D. program; this typically happens with a few students each year.
Ph.D. students typically serve as TAs during the first year, an experience that, along with a number of in-house presentations, is designed to expose students to the demands and excitement of teaching and course administration. Class presentations for both Ph.D. and MS students provide our students with valuable opportunities for learning the important skills of presentation and group interaction. Many of our graduate courses are project-based, which teaches all students the skills of identifying an open-ended problem, developing a solution, and implementing and testing that solution to determine the quality of its outcome. Many courses that implement projects also require in-class presentations about those projects.
Our Ph.D. students are required to meet breadth and depth requirements in preparation for a research career. The breadth requirement is met through a set of 10 graduate courses. Of the 10 courses, 2 must be completed in the systems area (from a set of 6 designated courses), and 2 in theory area (from a set of 6). At least half of the courses are typically completed during the student’s first year, while the remainder is completed before graduation. The depth requirement consists of a Ph.D. major area examination in the targeted specialization area and is usually completed in the second or third year of study. This consists of an oral presentation and a literature survey that demonstrate competence and preparation in the chosen specialty. The breadth and depth requirements are followed by a Ph.D. proposal and then a defense at the completion of research. Students usually start their research during the second year, but in some cases start during their first year.
The MS program offers three tracks: thesis, project or examination. Students in each track prepare a plan of study comprised of 6 courses from three areas: theory, systems, and applications. Students pursuing the examination option take an additional set of 3 courses and complete their degree with a comprehensive exam. Students pursuing the project option take a minimum of 6 units of directed research with a research advisor, write a project report and give a public presentation of their work. Students pursuing the thesis option complete a thesis under the advisement of a 3-member thesis committee and give a public defense of the thesis. The department also has a 5-year BS/MS program to encourage BS students to pursue a MS degree.
We offer a broad variety of graduate level courses and seminars that are a direct outcome of faculty specialties. We have regularly offered courses on many topics, including theory, databases, networks and wireless networks, security, programming languages, architecture, algorithms, graphics, computer vision, and machine learning, among others. We also offer “290” courses, which many be offered one time or multiple times, and are typically closely related to faculty research interests. Recent 290 offerings have included, e.g. Sparse Matrix Algorithms, Program Analysis, Advanced Data Mining, Networking for Multimedia, Green Computing, Querying and Indexing Contemporary Data, and Cryptographic Engineering. Finally, we also regularly offer “595” seminars. These are paper-centric, 2-unit courses (as opposed to 4-unit courses) that involve weekly meetings to discuss recent publications on a chosen topic. Students typically prepare presentations of the papers and participate in group discussions, giving students the opportunity to practice their presentation and technical reading skills.
b. Evaluate the department's success in attracting and retaining graduate students. How would one characterize the applicant pool at the graduate level in terms of quality, size, diversity, etc.? How is the program advertised? What procedures and criteria are used to select the students for admission? What percentage of graduate students admitted in the past five years have had an undergraduate GPA of under 3.0? Our graduate applicant pool has remained quite strong. At the PhD level, we typically have between about 250 and 300 applicants annually (with an unusual spike of 349 in 2007). The average PhD applicant has a GPA above 3.5 and quantitative GRE scores of 80% or better. At the MS level, we have between 200 and 250 applicants. A faculty committee makes admission decisions based on GPA, GRE scores, recommendation letters, statement of purpose and intended research area of the (PhD) student. We target a roughly steady size of student body in both the PhD and MS programs. We admit between 15% and 20% of PhD applicants; our “take rate” at the PhD level has grown from 34% in 2007 to 57% in 2009. At the MS level we admit roughly 20%-30% of applicants, with a take rate typically in the neighborhood of 30%. In the past five years we have admitted only one PhD applicant and two MS applicants with GPA below 3.0 (about 1.5% of admissions). Our applicant pool is strongly international, though the fraction of domestic applicants has trended generally upward in the past few years and now (2010) stands at 21% PhD / 17% MS. Our student body is also highly international (60% are non-resident aliens), with the largest international representation from China and India. Like the field of CS as a whole, our applicant pool is less diverse than we would like: In 2010, only ten of our graduate applicants were domestic URMs, and 21% of our applicants were female. Besides our web presence, the advertising we do for the graduate program is primarily targeted at increasing its diversity, as described under question (c) below.
c. What efforts are being made to recruit and retain students and to design the curriculum so as to achieve an appropriate ethnic and gender balance in the department? We have several strong efforts in the department to recruit and retain a diverse graduate student body. We have been particularly successful in the area of gender balance (considered against the background of the field of CS): In 2009-2010 our graduate student population is 23% female (43 of 147 students). Our Ph.D. student body is 26.4% female, compared to the 2007-08 national average of 19.3% female (Taulbee data). As part of our FireUP program, CS faculty visited and made a presentation about graduate school in general and UCSB graduate research in particular at specific CSU and UC campuses whose URM's were likely to consider UCSB because of its geographical proximity to their undergraduate campuses (Cal Poly, SLO,CSU Channel Islands, UC Merced, CSULA, UCSD). In addition, UCSB hosted a Graduate Preview Day, inviting URM seniors from nearby universities, organizing informational meetings, graduate student panels, and one-on-one meetings with faculty members. All three seniors who attended for computer science applied for fall 2010 admission. One was accepted. Once a student is accepted, we match the student with an existing graduate student with something in common (undergraduate institution, gender, ethnicity, home state, etc.) for personalized recruiting. Female students are matched with two students - one female and a second with something else in common. This allows females to ask questions they would not be comfortable asking a male (such as campus safety). The CSE IGERT and ICB Summer Applied Biotechnology Research Experience programs, headed by CS Prof. Petzold, are developing and strengthening our relationship with Jackson State University and are currently working on similar relationships with Florida Agricultural and Mechanical University, CSU Channel Islands, and UC Merced. In 2008, CS faculty and staff visited Jackson State University for the Louis Stokes Mississippi Alliance for Minority Participation (LSMAMP) Annual Symposium, where Prof. Petzold spoke about her work, the two programs, and the graduate research being done at UCSB. Female CS students are retained through an active club, WiCS (Women in Computer Science). It holds a minimum of one event each quarter, beginning with a kickoff at a female faculty or staff member's house. WiCS serves as both an academic and social networking group for our students. We do not have any graduate curriculum design efforts targeted specifically at diversity.
d. How is the quality of student performance assessed at various stages of students’ graduate careers? Do graduate students’ grades show a sufficient range? Do they do so early enough to identify students who lack the appropriate preparation and motivation? For PhD students the department has instituted a progress check every year. Both the students and their advisors are notified of the expected versus the actual progress of each student towards the degree, which is determined by the Graduate Affairs Committee. At the end of each academic year we have a faculty meeting whose sole purpose is the presentation and the discussion of the status of each and every one of our PhD students. Each student receives a letter indicating the outcome of the evaluation and expected next milestones, and any students receiving a “concerned” or “unsatisfactory” evaluation must meet with the graduate advisor to discuss their progress. The evaluation process keeps students abreast of the requirements and the expected normative time in which to complete them. There are also requirements and a number of milestones for the degree. The students have to pass an oral Major area exam. Following this, they are required to take a proposal exam and a final thesis defense. Regular meetings with research advisors occur throughout graduate work.
Grades in the range A+ through F grades are given, although the majority of the grades fall into the B- to A+ range. Ph.D. students must maintain a GPA of 3.5. Both Ph.D. students must earn at least a B in each course for that course to count towards the course requirement. MS students must earn at least a B in each of their four major area courses. A student is put on probation if their GPA drops below a 3.0, so this prompts action by the department. The MS program usually takes less than two years, and early feedback is based on the GPA.
e. What is the attrition/success rate (in terms of advancement to candidacy and graduation) for the department? What factors contribute to the department’s attrition/success rate? For example, how does the rate compare between those who entered with a 3.0 GPA or higher and those who entered with a GPA under 3.0? Are there differences in rates related to gender and ethnicity? Approximately 18% of our PhD students leave without completing a PhD. At least half of these students leave with a MS, while the remainder leave without a degree. While we do not keep statistics on the MS students, anecdotally we know it is much rarer that an MS student leaves without a degree. We believe the biggest factor that contributes to our Ph.D. program's success rate is our annual PhD progress checks, which ensure that each student is on-track. Any student that is flagged as “concerned” or “unsatisfactory” must meet with the graduate advisor to discuss a remediation plan. Assignment of a faculty advisor to each entering student, and the accessibility of our staff support, also contributes to our graduation rate. Factors that typically cause a PhD student to leave without a degree include two-body problems, change in career goals, and family concerns. In the last 5 years, the department admitted no MS students and only 1 PhD student with a GPA below a 3.0. This first year PhD student seems to be doing fine in her coursework and has started on research with an advisor. We are not aware of any differences in the rates related to gender and ethnicity. This could be because we do not explicitly track these statistics. However if we consider the students who failed to complete the PhD, there does not seem to be any correlation to gender or ethnicity. We note that the department has had 1% or less of underrepresented students in the last 5 years.
f. Evaluate the adequacy of support for graduate students. What percentage of students are funded, and how competitive are department financial offers with offers from other institutions? How are financial aid and fellowship decisions reached in the department? 100% of our PhD students are funded. We typically admit PhD students with a five-year support offer. The financial packages offered to PhD students may consist of campus fellowships (merit and diversity), and/or college and departmental fellowships. Some PhD students are supported as RAs beginning in their first year; and the remaining students (the majority of all incoming students) receive TAships with fellowships to cover nonresident tuition. PhD students begin their research no later than the second year, and most become RAs during their second year. While our support packages are generally adequate for living in Santa Barbara, the packages are less competitive than those that are offered by other universities. In particular, the raw dollar amount is typically less, and with a higher cost of living, students often indicate that they are concerned about the financial offer and may tend to favor other offers based on financial considerations. As a result, we believe that we are unable to attract some top students due to the inability to compete financially. We often try to supplement financial offers to top students through money raised from donors, and this has helped in our recruiting success. Ideally, we would like to be able to offer more competitive financial packages. We do not guarantee funding to our MS students, but the department currently supports just under half of its MS students as TAs or RAs; a number of others find jobs on campus in other departments. The Graduate Admissions Committee makes financial aid and fellowship decisions for newly admitted students, with input from the faculty regarding the availability of RAships.
g. How does the department provide academic advising? How does the department communicate the requirements of the program to incoming students? What other assistance does the department regularly provide to graduate students (e.g., tutorial help, study space, specialized libraries)? The department assigns an academic faculty advisor to every incoming student. The faculty member advises and helps the student prepare their study plan. A PhD student eventually finds a research advisor for advising, typically during the first or second year of study. New graduate students share a common computer lab. Once involved in research, students move to specific faculty research labs where each student typically receives their own desk, computer, and storage space. The program requirements are communicated to incoming students through a Fall orientation, and are always available online. We also provide students with a Graduate Student Handbook, containing all rules and guidelines. During the Fall orientation, one or more faculty members and the graduate program assistants make presentations that overview PhD and MS program requirements, and provide guidance for how best to complete those requirements. Information about faculty research is conveyed through a seminar series in the Winter quarter every year, where each faculty member makes a presentation about his/her research area.
h. What is the average time-to-degree for master’s students? Doctoral students? How do these averages compare with averages for similar programs in the UC or at other major institutions of higher education? How does the department ensure that students progress toward completion in a timely and orderly manner? The average time-to-degree for master’s students is 1.75 years (as of 2008-09). For the PhD students, this number is 5.75 years (as of 2008-09), and our normative time is 5 years. We are roughly the same in time to degree for the MS and PhD degrees within the UCSB College of Engineering departments, and about the same as the campus average for the MS degree. The department's time to PhD is one year less than the campus average. We ensure PhD students make timely progress towards graduation through the annual review of all PhD students. During this evaluation we check their progress in coursework as well as in research. If a student is found to be making insufficient progress, steps can be taken to get them back on track. We believe that in this way potential problems can be identified early on and prevented from harming the student’s progress. MS student progress is evaluated through their submission of an MS study plan. We also have a published time line (web site and grad student handbook) for PhD students where departmental expectations are clearly stated.
i. Describe the climate for graduate study that the department fosters (a) in general, (b) for women, and (c) for historically underrepresented students? If there are problems, what specific efforts are being made to address them? We believe that the climate for graduate study for all groups of students is conducive to good research and good student morale; we believe that the student survey results back up this claim. There are frequent student-faculty interactions inside and outside the classrooms. The department has regular coffee hours, where students and faculty interact in an informal setting. We also have 2 annual faculty/graduate student BBQs and a holiday party. Once a year, typically in the fall, our students organize and run a graduate student workshop, where they solicit submissions from their fellow students and a student program committee evaluates the submissions. There is strong industry involvement in the workshop. Twice yearly the chair has an open meeting with the graduate students to hear any compliments, complaints, or suggestions for improvement.
We have made it a priority to improve the climate for women in our department through the concerted hiring of female faculty (4 in 10 years) and recruiting of female graduate students. In the last 5 years, we have seen the percentage of female graduate students grow from 19% to 23%. With this growth we have established a graduate Women in Computer Science (WiCS) student group, which hosts frequent professional and social activities. Our numbers for under-represented minority (URM) students are, unfortunately, much lower, only 3% of our graduate population. We have been working to recruit more students in this category through securing funds from Graduate Division, through building recruiting partnerships with key CSU and UC campuses, and through applying for a DoE GAANN award.
j. What assistance is available in career planning and job placement? Comment on the department's placement record for graduate students who have received their degrees in the past five years. Academia day and connections to companies for industry jobs are the foremost efforts by the department. Our Academia day is a day of presentations, panels and events highlighting academia as a professional choice. Post-doctoral positions are also discussed, and interviewing tips are given. Our contacts with major companies (Google, Microsoft, Citrix, etc) and the positive record of our former students result in many companies actively recruiting on campus. Job placement is also done through the faculty's own contacts with companies. We do not track MS students like we do the PhD, although it is safe to say the majority of MS students get a job in industry. Our top students often join our PhD program or, less frequently, apply to PhD programs at other universities. Very few leave with no job or further graduate study plans.
For PhDs, the positions they obtained when they left UCSB are as follows: