Scott Fahlman
Research Faculty Emeritus - Contingent Carnegie Mellon University
- Pittsburgh PA
Scott Fahlman has worked in many areas of Artificial Intelligence, including the use of massively parallel machines to solve AI problems.
Biography
As a researcher, Scott is primarily interested in Artificial Intelligence and its applications. He has worked in many areas of AI: planning, knowledge representation and reasoning, image processing, natural language processing, document classification, artificial neural networks, and the use of massively parallel machines to solve AI problems. Scott is also interested in the use of AI techniques to build better user interfaces and context-aware systems.
Currently, Scott is working on Scone, a practical Knowledge Base System (KBS) that can represent a large body of real-world knowledge and that can efficiently perform the kinds of search and inference that seem so effortless for us humans. This work is based in part on the NETL system that Scott developed for his Ph.D. thesis in the late 1970s, but the Scone system is designed to run on standard workstations and servers rather than on special parallel hardware.
Scott's research group has worked on a number of applications of Scone, with a special focus on using Scone to support knowledge-based natural language understanding and generation. Scott believes that Scone-like knowledge base systems will be important tools in the future, perhaps used in even more ways than database systems are used today.
Scott is also working on some ideas for new learning architectures for deep-learning networks, inspired in part by the Cascade Correlation architecture that he developed in 1990 with Chris Lebiere.
Scott is a Fellow of the Association for the Advancement of Artificial Intelligence (AAAI).
Scott was one of the core developers of the Common Lisp language, and his research group developed the CMU Common Lisp implementation which formed the basis for many commercial Common Lisp systems, and now is maintained as open-source software, along with a split-off version, Steel Bank Common Lisp.
In 1982, Scott proposed the use of :-) and :-( in posts and Email messages. These are generally regarded as the first internet emoticons, and the text-only ancestors of today’s graphical emojis.
Areas of Expertise
Media Appearances
Today, the Emoticon Turns 30 :-)
The Atlantic online
2012-09-19
Today, at 11:44 a.m. Eastern Daylight Time, the emoticon will likely experience a jolt of joy. Today, at 11:44 a.m., the emoticon will likely experience a pang of angst. Today, at 11:44 a.m., the emoticon will likely take a moment to consider where it is, where it's been, and where it's going. Today, at 11:44 a.m., the emoticon will turn 30.
The emoticon was invented 33 years ago today — here's the professor who created it
Business Insider online
2015-09-19
Thirty-three years ago today, a professor at Carnegie Mellon University invented the emoticon. Scott E. Fahlman, along with other members of CMU's computer science community, used online "bulletin boards" to share information, make announcements, and chat, Fahlman recalled in a post on Carnegie Mellon's website.
The 40-year evolution from :-) to
CNN Business online
2022-09-18
At 11:44 a.m. on September 19, 1982, Scott Fahlman made internet history by stitching together a colon, a hyphen and a close parenthesis. Fahlman, a computer science professor at Carnegie Mellon University, posted “: - )” on the school’s online bulletin board, a primitive kind of social network accessible only by others on the university’s closed intranet and limited to text only.
My 24 hours without Facebook, Google, Apple, and Amazon
Yahoo Finance online
2019-08-01
It would be simple just to go off into the woods for a day and not engage with these companies, but that wasn’t the point. I wanted to do a full day of work (and some play). And of course some people are more dependent on the tech giants than others, which is usually a function of your preferences and what you do for a living. I think I’m somewhat representative of a typical professional. As Scott Fahlman, professor emeritus of computer science and artificial intelligence at Carnegie Mellon University says, “I could not unplug very long and keep doing my job.”
Want to Feel Old? the Emoticon Is 40 :-)
Gizmodo online
2022-09-19
The opportunities were endless as the early Internet stepped into its new role as a digital terra incognita. With a new frontier for humanity to explore and settle, one of the largest unknowns was how language would evolve in this new space. Given the missing opportunity to express emotion with visual cues, it would be too easy for meaning and humanity to be lost within the machinations of digital communication. So in an effort to communicate humor (and lack thereof) more efficiently, professor Scott E. Fahlman from Carnegie Mellon University proposed a pictographic hierarchy to delineate attempts at jokes on the university’s online bulletin board (called a “bboard”).
Media
Social
Industry Expertise
Accomplishments
Outstanding Technology Contributions Award
2013
Web Intelligence Consortium
Education
Massachusetts Institute of Technology
B.S.
Electrical Engineering and Computer Science
1973
Massachusetts Institute of Technology
Ph.D.
Artificial Intelligence
1977
Massachusetts Institute of Technology
M.S.
Electrical Engineering and Computer Science
1973
Affiliations
- American Association for Artificial Intelligence (AAAI) : Fellow
- Association for Computing Machinery : Member
Links
Event Appearances
Scientific and Artistic Creativity
Regional Arts Education Day Arts Education Collaborative, Pittsburgh, PA
Patents
Method for transforming message containing sensitive information
US5960080A
1999-09-28
A method for transforming an original message into a final message by including an untrusted service, includes the steps of identifying at least one sensitive term from the original message; replacing the at least one sensitive term with a standard token to create a sanitized message; storing the at least one sensitive term; transmitting the sanitized message to a provider of the untrusted service; performing the untrusted service on the sanitized message to create a serviced message; merging the serviced message with the at least one sensitive term stored in the storing step to create the final message.
Articles
A semantic model for actions and events in ambient intelligence
Engineering Applications of Artificial Intelligence2011
Event ma nagement and response generation are two essential aspects of systems for ambient intelligence. This work proposes handling these issues through the use of a semantic model for ambient intelligence which, under the umbrella of a philosophical and common-sense optic, describes what actions and events are, how they are connected, and how computational systems should think about their meaning.
Possible-World and Multiple-Context Semantics for Common-Sense Action Planning
Situational Awareness for Assistive Technologies2012
Event management and response generation are two essential aspects of systems for Ambient Intelligence. In this regard, the context notion does also play an essential role, not only in determining the set of activities that take place in it, but also in devising the most appropriate response to those situations. Context is also essential for disambiguating knowledge and meaning.
Beyond Idiot-Savant AI
Advances in Cognitive Systems2012
Why has progress toward broad, flexible, human-like AI been so slow? I suggest the key reason is that, in recent years, few researchers have actually focused on this goal. Instead they have focused on achieving super-human (but brittle) performance in a few narrow problem domains. That is a valuable enterprise, but some attention should also be paid to the original – and still unachieved – goal stated decades ago by the founders of the AI field.
Image Retrieval with Textual Label Similarity Features
Intelligent Systems in Accounting, Finance and Management2015
This article presents a knowledge-based solution for retrieving English descriptions of images. We analyse the errors made by a baseline system that relies on term frequency, and we find that the task requires deeper semantic representation. Our solution is to perform incremental, task-driven development of an ontology. Ontological features are then applied in a machine-learning algorithm for ranking candidate image descriptions.
Sensing, Perceiving, and Understanding Actions
International Journal of Distributed Sensor Networks2014
With the emergence of paradigms such as the smart city, Internet of Things, or the smart grid among some of the most relevant, distributed sensor networks have gained importance as an enabling tool for information gathering. However, in order to leverage these paradigms, the distributed systems are required to be endowed with advanced reasoning capabilities, whereas the retrieval of contextual information or the application of a set of simple rules does not suffice since elaborated responses are expected from these systems.
