Encyclopedia of Computer Science and Technology

190 Feigenbaum, Edwardhttp://hissa.nist.gov/chissa/SEI_Framework/framework_1.html. Accessed September 20, 2007.Pullum, Laura L. Software Fault Tolerance: Techniques and Implementation.Norwood, Mass.: Artech House, 2001.Feigenbaum, Edward(1936– )AmericanComputer ScientistEdward Feigenbaum is a pioneer artificial intelligenceresearcher, best known for his development of expert systems(see artificial intelligence). Feigenbaum was bornin Weehawken, New Jersey. His father, a Polish immigrant,died before Feigenbaum’s first birthday. His stepfather, anaccountant and bakery manager, was fascinated by scienceand regularly brought young Edward to the Hayden Planetarium’sshows and to every department of the vast Museumof Natural History. The electromechanical calculator hisfather used to keep accounts at the bakery particularly fascinatedEdward. His interest in science gradually turned toa perhaps more practical interest in electrical engineering.While at the Carnegie Institute of Technology (nowCarnegie Mellon University), Feigenbaum was encouragedto venture beyond the more mundane curriculum tothe emerging field of computation. He became interestedin John Von Neumann’s work in game theory and decisionmaking and also met Herbert Simon, who was conductingpioneering research into how organizations madedecisions (see von Neumann, John). This in turn broughtFeigenbaum into the early ferment of artificial intelligenceresearch in the mid-1950s. Simon and Alan Newell had justdeveloped Logic Theorist, a program that simulated theprocess by which mathematicians proved theorems throughthe application of heuristics, or strategies for breaking problemsdown into simpler components from which a chain ofassertions could be assembled leading to a proof.Feigenbaum quickly learned to program IBM mainframesand then began writing AI programs. For his doctoralthesis, he explored the relation of artificial problemsolving to the operation of the human mind. He wrote acomputer program that could simulate the human processof perceiving, memorizing, and organizing data forretrieval. Feigenbaum’s program, the Elementary Perceiverand Memorizer (EPAM), was a seminal contribution to AI.Its “discrimination net,” which attempted to distinguishbetween different stimuli by retaining key bits of information,would eventually evolve into the neural network (seeneural network). Together with Julian Feldman, Feigenbaumedited the 1962 book Computers and Thought, whichsummarized both the remarkable progress and perplexingdifficulties encountered during the field’s first decade.During the 1960s, Feigenbaum worked to develop systemsthat could perform induction (that is, derive generalprinciples based on the accumulation of data about specificcases). Working on a project to develop a mass spectrometerfor a Mars probe, Feigenbaum and his fellow researchersbecame frustrated at the computer’s lack of knowledgeabout basic rules of chemistry. Feigenbaum then decidedthat such rules (or knowledge) might be encoded in sucha way that the program could apply it to the data beinggathered from chemical samples. The result in 1965 wasDendral, the first of what would become a host of successfuland productive expert systems (see expert system). Afurther advance came in 1970 with Meta-Dendral, a programthat could not only apply existing rules to determinethe structure of a compound, it could also compare knownstructures with the existing database of rules and infer newrules, thus improving its own performance.During the 1980s, Feigenbaum coedited the four-volumeHandbook of Artificial Intelligence. He also introduced expertsystems to a lay audience in two books, The Fifth Generation(co-authored with Pamela McCorduck) and The Rise of theExpert Company.Feigenbaum combined scientific creativity with entrepreneurshipin founding a company called IntelliGeneticsand serving as a director of Teknowledge and IntelliCorp.These companies pioneered the commercialization ofexpert systems. In doing so, Feigenbaum and his colleaguespublicized the discipline of “knowledge engineering”—thecapturing and encoding of professional knowledge in medicine,chemistry, engineering, and other fields so that it canbe used by an expert system. In what he calls the “knowledgeprinciple” he asserts that the quality of knowledge ina system is more important than the algorithms used forreasoning. Thus, Feigenbaum has tried to develop knowledgebases that might be maintained and shared as easily asconventional databases.Remaining active in the 1990s, Feigenbaum was secondpresident of the American Association for Artificial Intelligenceand (from 1994 to 1997) chief scientist of the U.S.Air Force. In 1995, Feigenbaum received the Association forComputing Machinery’s prestigious A. M. Turing Award.Founder of the Knowledge Systems Laboratory at StanfordUniversity, Feigenbaum remains a professor emeritus ofcomputer science at that institution.Further ReadingFeigenbaum, Edward, Julian Feldman, and Paul Armer, eds. Computersand Thought. Cambridge, Mass.: MIT Press, 1995.Feigenbaum, Edward, Pamela McCorduck, and H. Penny Nii. TheRise of the Expert Company: How Visionary Companies areUsing Artificial Intelligence to Achieve Higher Productivity andProfits. New York: Vintage Books, 1989.Henderson, Harry. Artificial Intelligence: Mirrors for the Mind. NewYork: Chelsea House, 2007.Shasha, Dennis, and Cathy Lazere. Out of Their Minds: The Livesand Discoveries of 15 Great Computer Scientists. New York:Copernicus/Springer-Verlag, 1995.fiber opticsA fiber optic (or optical fiber) cable transmits photons(light) instead of electrons. Depending on the diameter ofthe cable, the light is guided either by total internal reflectionor as a waveguide (manipulating refraction). Theseprinciples were known as early as the mid-19th century andbegan to be used in the 20th century for such applications