Computer science pioneer David L Waltz dies aged 68

David L. Waltz, a computer scientist whose early research in information retrieval provided the foundation for today’s Internet search engines, died on Thursday in Princeton, N.J. He was 68. The cause was brain cancer, his wife, Bonnie Waltz, said. He died at the University Medical Center at Princeton. During his career as a teacher and a technologist at start-up companies as well as large corporate laboratories, Dr. Waltz made fundamental contributions to computer science in areas ranging from computer vision to machine learning. One signal achievement was the development of a basic technique that makes it possible for computers to render three-dimensional scenes accurately. As part of his Ph.D.   dissertation at the Massachusetts Institute of Technology, he developed an algorithm that could extract a rich three-dimensional understanding of a scene from two-dimensional line drawings with shadows. The 3-D research was seminal in the fields of computer vision and artificial intelligence. Known as “constraint propagation,” the technique is now used in industry for solving problems like route scheduling, package routing and construction scheduling. At M.I.T., Dr. Waltz was taught by Marvin Minsky, a pioneer in artificial intelligence. Dr. Waltz graduated in 1972, then taught computer science at the University of Illinois at Urbana-Champaign and, later, at Brandeis University in Massachusetts. But it was as a member of a group of researchers at the Thinking Machines Corporation, in Cambridge, Mass., that Dr. Waltz made his breakthrough in information retrieval. Thinking Machines was an early maker of massive, parallel supercomputers, and by joining the company, in 1984, Dr. Waltz gained access to computers that by ’80s standards held vast amounts of fast random-access memory, up to 512 megabytes. “For the first time it was possible to use simple algorithms with lots and lots of data,” said Brewster Kahle, a computer scientist who directs the Internet Archives and was one of the Thinking Machines researchers. Access to that database was crucial to Dr. Waltz’s development of a technique known as memory, or “case based,” reasoning. It revolutionized the way computers recognized characters, words, images and later, even voices. Before, a computer had to follow a set of programmed rules to arrive at recognition (it’s an “i” if there’s a dot, for example). Now it could comb through its vast memory and deduce what the image was by comparing it to what had been stored there. The technique transformed the field of artificial intelligence and also greatly advanced voice recognition and machine vision technology. And it led directly to the “big data” and data-science approaches that are essential tools for search engines, allowing them to sift through large collections of information to improve accuracy and relevance. “He was a real pioneer,” said Peter Norvig, Google’s director of research. “The two main changes that got us modern A.I. were probabilistic reasoning and using memory rather than rules.” “I don’t know if Larry and Sergey read his papers directly,” he added, referring to Google’s founders, Larry Page and Sergey Brin, “but the idea, filtered through however many people, was certainly a key.” While at the University of Illinois, Dr. Waltz turned to the field of natural language understanding, a component of artificial intelligence involving the interpretation of language. With support from the Office of Naval Research, he built a question-answering system called Planes and explored the use of neural networks in language processing. In another early project, a Thinking Machines group led by Dr. Waltz designed an information retrieval system that made it possible for a remote user to gain access to a supercomputer and then be able to search through large volumes of documents. The system, known as Wide Area Information Server, or WAIS, and designed in cooperation with the Dow Jones Corporation, Apple Computer and KPMG Peat Marwick, was not the first information retrieval system. But it was innovative in enabling the user to uncover connections between seemingly disparate documents. For example, the WAIS system was able to give an early warning of the Chernobyl disaster in 1986 after it discovered a report of an abnormal radiation reading in Scandinavia, according to W. Daniel Hillis, the co-founder of Thinking Machines. WAIS also introduced techniques to narrow a document search. It was followed by other search systems, like Veronica, Gopher and Archie, which predated the search engines offered today by Google, Microsoft, Yahoo and other companies. After leaving Thinking Machines in 1993, Dr. Waltz joined the NEC Research Institute in Princeton, where he was president from 2000 to 2002. He left to help create the Center for Computational Learning Systems at Columbia, where he was director. The center has worked with Con Edison of New York in developing systems that can predict power failures and thus enhance maintenance of the electric power grid. Researchers there are also working on creating a computer-based system to give people with epilepsy early warnings of seizures. The technique involves mining data generated by electrodes implanted in patients. Dr. Waltz earlier was instrumental in establishing interdisciplinary research centers: the Beckman Institute at the University of Illinois, and the Volen National Center for Complex Systems at Brandeis. David Leigh Waltz was born in Boston on May 28, 1943, to Maynard C. Waltz and the former Lubov Leonovich. His father, a physicist, worked at M.I.T.’s Radiation Laboratory during World War II and later at Bell Labs. Dr. Waltz obtained both undergraduate and graduate degrees at M.I.T. in electrical engineering. He lived in Princeton. Besides his wife, he is survived by a brother, Peter; a son, Jeremy; a daughter, Vanessa Waltz, and a granddaughter.