[Published originally in the September 2002 edition of Computing Research News, Vol. 14/No. 4, p. 4.]
Computer Science Research in Mexico
By Valerie Bernat
The current status of computer science research in Mexico is of interest to researchers in the United States not only for the intrinsic value of the science, but also because the methods Mexico is using to attempt a quantum leap forward in scientific capability can be seen as a model for other countries in the developing world.
Because the research structure in Mexico is unfamiliar to many in the United States, we will first look at the players, the sources of funding, and the constraints on research before considering the efforts to improve the quality and quantity of computer science research in Mexico.
CONACyT (National Council of Science and Technology) is the counterpart of NSF in Mexico, providing funding for all areas of science and technology.
The top public universities, IPN (National Polytechnic Institute) and UNAM (National Autonomous University of Mexico), receive operational funding from the federal government. Salaries are low, about U.S. $28,000 annually, and faculty teach one to two courses per semester. Tuition for students is quite low. When UNAM recently proposed an increase from $0.02 to $145 per year, students protested with a loud and lengthy strike.
IPN recently founded a new research center that has grown rapidly with special institutional support for better salaries via research grants from CONACyT. Some scientists from Russia and Cuba are participating in this effort.
The top private universities for research are ITESM (Institute of Technology and Higher Studies of Monterrey: Monterrey, Morelos, and State of Mexico campuses) and UDLA (University of the Americas: Puebla campus). Private universities are funded through foundations, usually supplemented with resources from industry and tuition. Until recently, due to Mexico's egalitarian attitudes with regard to public monies, private institutions were unable to receive government research support. Salaries are about U.S. $36,000 annually, and faculty teach 3 courses per semester.
Separate from universities, the Mexican federal government invests in a number of public scientific research institutes, several with international reputations. A board and a director oversee each center and choose areas of research to pursue within funding allocations. In the area of computing, CIMAT (Center for Investigation in Mathematics) specializes in image processing, CICSE (Center for Scientific Investigation and Higher Education of Ensenada) focuses on software engineering and cluster programming, and INAOE (National Institute of Astrophysics, Optics and Electronics) works in computer control.
There is one independent computing research institute in Mexico. In 1991 a group of UNAM faculty established LANIA (National Laboratory of Advanced Computer Science) so that they could pursue research in theoretical artificial intelligence, including multi-agent systems and logic. With additional funding opportunities and government requests, LANIA has grown and now includes research in the areas of image processing and computer vision, programming languages and methodologies, distributed and cooperative computing, general IT consulting, human resources training, and direct government support such as computer network installation.
Almost all research in computing is done in these academic institutions and research centers. Industry has made some efforts to establish research facilities, but the lack of qualified researchers has hampered their efforts.
Through 2000, approximately 160 Ph.D. computer scientists were working in Mexico; most received their degrees from universities in the United States, France, England, Japan, and Spain. In recent years, ITESM, UNAM, UDLA-P, and IPN have established Ph.D. programs.
Constraints, Funding, and the Current Scene
Although the first supercomputer in Latin America was installed at UNAM, in the past Mexico did not see computing as a strategic investment. Funds were available for computing only as it became useful in pursuing other scientific research. CS as a discrete discipline was not viewed as worthy of consideration or support by the scientists running CONACyT, whose motivation may have also been partly a reluctance to share limited resources with an upstart discipline that could be construed as more technical than truly scientific.
In addition, the limitations of salaries and the value of the peso have constrained the productivity of researchers in Mexico, making it more difficult for faculty and students to travel, join international associations, subscribe to journals, or purchase texts or state-of-the-art equipment. Unlike their U.S. counterparts, faculty in Mexico do not receive sample copies of textbooks.
Prior to 1994, CONACyT had two programs affecting computer science research: funding for graduate students and the National Investigators Program. Students seeking a Ph.D. at a qualified institution in Mexico, the United States, or overseas would receive a loan that would be entirely forgivable if the student returned to an academic career within Mexico. The loan would be half forgiven if the student returned to take a job in industry. Approximately 90 percent have completed their degrees, and virtually all have returned to work in Mexico.
The National Investigators Program was established to compensate for low academic salaries in the scientific and technical disciplines by providing additional support. This program awards bonuses of up to 100 percent of salary, tax-free. The sole criterion for receiving an award is by number of pure research papers in international journals. Panels of distinguished researchers control the process. As a relative newcomer to the pantheon of disciplines within Mexico, computer science faculty have had difficulty gaining a voice and, thereby, a share of this funding. The panels have tended to favor their own research areas, disciplines such as astronomy and anthropology, with long, distinguished traditions in Mexico. In addition, no accommodation is made for the varying definitions of research or for the varying difficulties of producing and publishing pure research among the different disciplines. Unable to make progress without support, computer scientists did not meet the qualifications for support, a classic catch-22.
CONACyT first supported computer science in an effort with NSF conceived of by A. Nico Habermann, then AD for CISE. Following high-level meetings, Habermann pursued research collaboration and cooperation with CONACyT in CISE fields. Oscar Garcia, then a Program Director in CISE, continued the effort after Habermann's death, and funded the first of a series of three joint U.S./Mexico workshops focusing on collaboration between researchers in the United States and Mexico.
As a result of the first of these workshops, SMCC (Mexican Society for Computer Science) was formed. In addition, by joining NSF in funding this workshop and the joint research program that resulted, CONACyT was for the first time providing funds directly to computer science. NSF and CONACyT continue to support joint research projects, and their success, along with the example of NSF interest in computer science, has created a revolution in attitudes within CONACyT. Since 1998, programs initiated by CONACyT have begun to put computer science research on a strong footing within the country and the international community.
Working at Solutions
REDII (Network for Development and Investigation in Informatics) was established by CONACyT in 1998 with the goal of transforming computer science research and education within Mexico. Recognizing that Mexico cannot build a research capacity in computer science without also developing the people who are going to do the research--and knowing that in a developing country pure research must be balanced with an emphasis on practical development--REDII had a dual intent: to simultaneously fund interesting, useful research, while also supporting the future generation of researchers.
REDII funded both public and private institutions, not individuals. The institutions then supported projects or general development, but the monies could not be used for salary or release time. REDII was unique in that it was set up to directly fund research in computer science, not the panoply of scientific disciplines within Mexico. And, proposals were reviewed by CS research scientists. Of the nine institutions and more than 40 projects currently funded, two examples are Enciclomedia and Phronesis.
Enciclomedia uses research in HCI to couple textbooks with a multimedia database, creating a user-centered, customized learning tool with limitless possibilities. Researchers are currently engaged in producing a prototype to be used in the 5th grade. Through Enciclomedia, Mexico is actively developing the reason to link schools to the internet in the future, anticipating the time when the effort will be cost effective (cetee.itam.mx/redii/informe2/ProyectosPrototipos/Proyectos2.htm).
Phronesis is a web-based system for building and using distributed digital library collections. Phronesis provides space and time efficient procedures for indexing, searching and retrieving information in either English or Spanish (copernico.mty.itesm.mx/~tempo/Projects/).
The current administration is devising a follow-on program to REDII.
Mexico has made tremendous improvement in the last 10 years in increasing the quality and quantity of computer science research. Particularly worthy of note is the philosophy of looking ahead by investing in computer science research, though the vicissitudes of governmental focus and funding can still leave the field in a precarious position. In Mexico, as throughout the world, the need for computer capability outpaces the supply of qualified computer science professionals. Computer science is not a field that can be ignored.
Valerie Bernat is a freelance writer who lives in the Washington, DC, area.
This is one of an occasional series of articles describing computing research in other countries.
Copyright © 2007 Computing Research Association. All Rights Reserved. Questions? E-mail: email@example.com.