Curriculum Vitae: Scott E.
Fahlman
Research Professor RESEARCH INTERESTS
Artificial Intelligence and its applications, especially knowledge representation, planning, natural language, and machine learning. I have also worked in artificial neural networks, image processing, and document management. Languages and programming environments for evolutionary development of complex software systems, especially languages of the Lisp family. EDUCATION
Ph.D. in Artificial Intelligence, Massachusetts Institute of Technology, 1977. Dissertation in the M.I.T. Artificial Intelligence Laboratory under Professor Gerald J. Sussman: "NETL: A System for Representing and Using Real-World Knowledge". M.S. in Electrical Engineering and Computer Science, Massachusetts Institute of Technology, 1973. Master's thesis under Professor Patrick H. Winston: "A Planning System for Robot Construction Tasks". B.S. in Electrical Engineering and Computer Science, Massachusetts Institute of Technology, 1973. EMPLOYMENT
July
1978-Present: July
2000- April 2003: Research Staff Member, May
1996-July 2000: September
1977–June 1978: N. E. Berg Company, Summers
1965-1972: Summer jobs while at M.I.T. were at Honeywell Computer
Science Center, Digital Equipment Corporation, CAREER HIGHLIGHTS
Developed the BUILD program (Master's thesis at MIT), one of the first planning programs to implement plan modification and failure-directed backtracking. Developed the NETL knowledge representation system (Ph.D. thesis at MIT), which stored real-world, common-sense knowledge in a massively parallel network of simple hardware elements and links. This "data-parallel" approach made it possible to perform human-like inference and retrieval tasks very quickly and flexibly. NETL was a precursor of data-parallel computers (as in the Connection Machine) and "connectionist" approaches to Artificial Intelligence. Member of the 5-person core design team for the Common Lisp programming language. Moderator for internet-based design discussions involving dozens of major contributors. Head of the CMU Common Lisp project that produced a widely used free implementation of Common Lisp. At its peak, CMU CL was used by at least 10 companies as the basis for their own Common Lisp implementations. The CMU CL compiler set a new standard of performance for languages of the Lisp family. Developed a number of learning algorithms for artificial neural networks, including Quickprop, Cascade-Correlation, and Recurrent Cascade-Correlation. These algorithms typically learn desired behaviors10 to 100 times faster than older algorithms. Principal Investigator for the "Automated Interactive Microscope" (AIM) project, whose goal was to couple high-performance computing to advanced light microscopes for real-time processing of biomedical images. Head of the Gwydion Project at Carnegie Mellon. This project worked with Apple Computer to define the Dylan programming language and developed two public-domain implementations of Dylan. We also developed an advanced software development environment for the Java language. As President and CTO of Just Research, created a well-regarded computer science research lab near the Carnegie Mellon campus. In 2 years, built up the lab from initial conception to 25 full-time researchers. Overall
manager and system architect during the first phase of the RADAR Project in
Carnegie Mellon’s Developed the Scone Knowledge-Base system, based in part on ideas from my earlier NETL work, but optimized to run well on standard high-end workstations. HONORS
Fellow of the American Association for Artificial Intelligence (AAAI). PUBLICATIONS
Fahlman, S. E. (1974) "A Planning System for Robot Construction Tasks", Artificial Intelligence 5 (1974), 1-49. Reprinted in Tutorial on Robotics, 2nd Edition, C. S. G. Lee, R. C. Gonzalez, and K. S. Fu (eds.), IEEE Computer Society Press, 1986. Fahlman, S.
E. (1979) NETL: A System for Representing and Using Real-World Knowledge, MIT
Press, Fahlman, S. E. (1980) "The Hashnet Interconnection Scheme", Technical Report, CMU Computer Science Department. Fahlman, S. E. (1980) "Design Sketch for a Million-Element NETL Machine", Proceedings of AAAI-80 Conference, Morgan-Kaufmann, Los Altos CA. Reprinted in Computers for AI Applications, B. Wah and G. J. Li (eds.), IEEE Computer Society Press, 1986. Fahlman, S.
E. (1981) "Representing Implicit Knowledge" in Parallel
Models of Associative Memory, G. E. Hinton and J. A. Anderson
(eds.), Erlbaum Associates, Fahlman, S.
E. (1981) "Computing Facilities for AI: A Survey of Present and
Near-Future Options", AI Magazine, March 1981. Reprinted in
Computers for AI Applications, B. Wah and G. J. Li
(eds.), IEEE Computer Society Press, 1986. Also reprinted in Fahlman, S. E., D. S. Touretzky, and W. van Roggen
(1981) "Cancellation in a Parallel Semantic Network"
Proceedings of IJCAI-81, Fahlman,
S. E. (1982) "Three Flavors of
Parallelism", Proceedings of the Fourth National Conference, Canadian
Society for Studies of Intelligence, Fahlman, S. E., G. E. Hinton, and T. J. Sejnowski (1983) "Massively Parallel Architectures for AI: NETL, Thistle, and Boltzmann Machines" in Proceedings of the AAAI-83 Conference, Morgan-Kaufmann, Los Altos CA. Reprinted in Computers for AI Applications, B. Wah and G. J. Li (eds.), IEEE Computer Society Press, 1986. Fahlman, S.
E. and S. P. Harbison (1984) "The Spice
Project" in Interactive Programming Environments, Major contributor to Steele, G. L. Jr. (1984) Common Lisp: The Language, Digital Press. Wholey, J. S. and
S. Fahlman, S. E. (1985) "Parallel Processing in Artificial Intelligence" (extended abstract) in Parallel Processing 2, North-Holland Publishers. Fahlman, S. E. (1988) "Parallel Processing in Artificial Intelligence" (full paper) in "Parallel Computation and Computers for Artificial Intelligence", J. S. Kowalik (editor), Kluwer Academic Publishers. Fahlman,
S. E. and G. Fahlman, S. E. (1987) "Common Lisp" in Annual Review of Computer Science, Volume 2, 1987, Annual Reviews, Inc., Palo Alto, CA. Fahlman,
S. E. (1988) "Faster-Learning
Variations on Back-Propagation: An Empirical Study" in Proceedings, 1988
Connectionist Models Summer School, Morgan-Kaufmann, Fahlman, S. E. (1989) "Distributed Connectionist Systems for AI: Prospects and Problems" by Scott E. Fahlman, in Concepts and Characteristics of Knowledge-Based Systems, M. Tokoro, Y. Anzai, and A.Yonezawa (eds.), North-Holland Publishers. Fahlman, S. E. (1990) "What Can Massively Parallel Architectures Bring to AI?" Proceedings of 2nd Australian Joint Artificial Intelligence Conference, C. J. Barter and M. J. Brooks (eds.), Springer Verlag (Lecture Notes in AI Series). Fahlman,
S. E. and C. Lebiere (1990) "The Cascade-Correlation Learning
Architecture" in Advances in Neural Information Processing Systems 2, D.
S. Touretzky (ed.), Morgan-Kaufmann, Fahlman, S. E. and D. B. McDonald (1990) "Design Considerations for CMU Common Lisp" by Scott E. Fahlman and David B. McDonald, in Peter Lee (ed.) Advanced Language Implementation Techniques, Springer-Verlag. Fahlman,
S. E. (1991) "The Recurrent
Cascade-Correlation Architecture" in Advances in Neural Information
Processing Systems 3, D. Hoehfeld, M. and S. Hoehfeld, M. and S. E. Fahlman (1992) "Probabilistic rounding in neural Network Learning with limited precision" in Neurocomputing 4 (1992) pp. 291-299, Elsevier Science Publishers. Wah, B. W., S. E. Fahlman, and 22 co-authors (1993), "Report on Workshop on High Performance Computing and Communications for Grand Challenge Applications: Computer Vision, Speech and Natural Language Processing, and Artificial Intelligence" in IEEE Transactions on Knowledge and Data Engineering, Vol. 5, No. 1, February 1993. Baluja, S. and S. E. Fahlman (1994), "Reducing
Network Depth in the Cascade-Correlation Learning Architecture", Tech
Report CMU-CS-94-209, Fahlman, S. E., G. L. Steele Jr., H. Shrobe, B. Stroustrup, and C. A. Knoblock (1996), "The Future of Lisp" in IEEE Expert, vol. 11, no. 1, pp. 10-16, Feb. 1996. Fahlman, S. E. (2002), “Selling interrupt rights: a way to control unwanted E-mail and telephone calls”, IBM Systems Journal, Volume 41, Number 4, 2002, pages 759-766, November 2002. Tribble, A.
and S. E. Fahlman (2006), "Resolving Noun Compounds with Multi-Use
Domain Knowledge", FLAIRS 2006 conference, Tribble, A.,
B. Lambert, and S.E. Fahlman (2006), "SconeEdit: a Text-Guided Domain Knowledge Editor", (demo)
HLT-NAACL 2006 Conference, Fahlman, S.
E.(2006), "Marker-Passing Inference in the Tribble, A.
and S.E. Fahlman (2007), "CMU-AT:
Semantic Distance and Background Knowledge for Identifying Semantic
Relations" Proceedings of SemEval-2007: 4th International Workshop on
Semantic Evaluations. Shultz, T.R. and S.E. Fahlman, S.E. (2010), “Cascade-Correlation”, chapter in Encyclopedia of Machine Learning, C. Sammut and G. Webb (eds.), Springer-Verlag. Santofimia, M. J., S.E. Fahlman, F. Moya, and J.C. López (2010), “A Common-Sense Planning Strategy for Ambient Intelligence”, 14th International Conference on Knowledge-Based and Intelligent Information and Engineering Systems (KES 2010), Cardiff, Wales, UK. Fahlman, S.
E. (2011), “Using Scone’s multiple-context
mechanism to emulate human-like reasoning”, Proceedings
of the AAAI Fall Symposium on Advances in Cognitive Systems. Fahlman,
S.E. (2012), “Beyond Idiot-Savant AI” in Advances
in Cognitive Systems 1, pages 15-22.
SOFTWARE SYSTEMS
CMU Common Lisp. A complete, high-performance, public-domain implementation of Common Lisp with compiler, Hemlock text editor, and extensive programming environment. Represents about 35 person-years of effort. Widely used public-domain implementations of the Quickprop, Cascade-Correlation, and Recurrent Cascade-Correlation learning algorithms for neural networks. The CMU Benchmark Collection of benchmark problems for neural-net learning algorithms. Gwydion Dylan. Two implementations of the Dylan programming language. "Mindy" is a very flexible and portable byte-coded implementation, but rather slow. The "d2c" compiler produces much faster code, but is less portable. Sheets. An advanced software development environment for Java based on the "hypercode" concept. In order to facilitate ongoing development of a software system, we must retain all kinds of knowledge about a software system, including code, structured documents, and the links between them. PATENTS:
US Patent #5960080, "Method for transforming message containing sensitive information", Scott E. Fahlman, Rahul Sukthankar, Antoine Brusseau. PROFESSIONAL ACTIVITIES: Member, DARPA Software Research Review Panel (1995). Associate editor for the journal Neural Computation (1993-1998). Member of the editorial board for the journal Neurocomputing (1996-1998). Member of the editorial board, AAAI Press (1990-1994). Member of the editorial board of IEEE Computer (1988-1992). Program co-chair for Algorithms and Architectures, 1993 and 1994 Neural Information Processing Systems Conference. Keynote
speaker twice at Australian National AI Conference: 1988 ( Invited participant, NSF Workshop on Grand Challenges in AI (Washington DC, 1992). Keynote speaker at 53rd Annual Conference of the Association of Business Professors (Linz, Austria, 1991). Invited panelist, Japanese Science Foundation, Science and Technology Forum '90 (Tsukuba, Japan, 1990). Keynote speaker at 4th IEEE Conference on AI Applications (San Diego, 1988). Invited
speaker, British Computer Society ( Invited participant/speaker, IFIP Workshop on Concepts and Characteristics of Knowledge-Based Systems (Gotemba, Japan, 1987). Workshop leader at 1989, 1990, 1991, 1992 NIPS conferences. Lecturer at 1986 and 1988 Connectionist Models Summer School. Member of
Scientific Panel for Basics of Information Technology, Siemens Corporation
Central Research Laboratories, Member of X3J13 committee for the official standardization of Common Lisp, 1986 - 1988. Member of the editorial board for Advances in Cognitive Systems online journal. Consultant (formerly) for Justsystem Corporation, Fostin Capital Associates, Digital Equipment Corporation, Lucid Inc, Data General, Xerox, IBM, Perq Systems, Expert Technologies Inc., Control Data Corporation, Carnegie Group, Texas Instruments, Topologix, Cable Howse & Cozadd, Warburg-Pincus Ventures. Frequent reviewer for NSF funding proposals, major AI journals and conferences. Member of the Association for Computing Machinery and the American Association for Artificial Intelligence. PERSONAL
U.S. citizen. Hobbies (when I have time) include digital photography, cooking, and writing. |