The Cyc project is predicated on the idea that effective machine learning depends on having a cor... more The Cyc project is predicated on the idea that effective machine learning depends on having a core of knowledge that provides a context for novel learned information – what is known informally as "common sense." Over the last twenty years, a sufficient core of common sense knowledge has been entered into Cyc to allow it to begin effectively and flexibly supporting its most important task: increasing its own store of world knowledge. In this paper, we present initial work on a method of using a combination of Cyc and the World Wide Web, accessed via Google, to assist in entering knowledge into Cyc. The long-term goal is automating the process of building a consistent, formalized representation of the world in the Cyc knowledge base via machine learning. We present preliminary results of this work and describe how we expect the knowledge acquisition process to become more accurate, faster, and more automated in the future.
The Cyc project is predicated on the idea that effective machine learning depends on having a cor... more The Cyc project is predicated on the idea that effective machine learning depends on having a core of knowledge that provides a context for novel learned information – what is known informally as "common sense." Over the last twenty years, a sufficient core of common sense knowledge has been entered into Cyc to allow it to begin effectively and flexibly supporting its most important task: increasing its own store of world knowledge. In this paper, we present initial work on a method of using a combination of Cyc and the World Wide Web, accessed via Google, to assist in entering knowledge into Cyc. The long-term goal is automating the process of building a consistent, formalized representation of the world in the Cyc knowledge base via machine learning. We present preliminary results of this work and describe how we expect the knowledge acquisition process to become more accurate, faster, and more automated in the future.
This paper presents a novel method, based on the Cyc Knowledge Base and Inference Engine, of gath... more This paper presents a novel method, based on the Cyc Knowledge Base and Inference Engine, of gathering, organizing and sharing information about entities of interest (be they people, organizations , events or some other type of entity). The formal representations used in the Fact Sheets allow users to easily share information with others , run automated queries against the information , and allow the system to attempt to automatically gather and verify information before presenting it to the analyst. The system automatically keeps track of provenance (both which document a fact came from, and who interpreted the document). When gathering information automatically, the system produces a variety of search strings (using all known names for the entity) and then scours its sources for possible answers. Individual analysts can specify what types of information they are interested in for different types of entities, and can also specify additional patterns that can be used for finding tha...
This paper presents a novel method, based on the Cyc Knowledge Base and Inference Engine, of gath... more This paper presents a novel method, based on the Cyc Knowledge Base and Inference Engine, of gathering, organizing and sharing information about entities of interest (be they people, organizations , events or some other type of entity). The formal representations used in the Fact Sheets allow users to easily share information with others , run automated queries against the information , and allow the system to attempt to automatically gather and verify information before presenting it to the analyst. The system automatically keeps track of provenance (both which document a fact came from, and who interpreted the document). When gathering information automatically, the system produces a variety of search strings (using all known names for the entity) and then scours its sources for possible answers. Individual analysts can specify what types of information they are interested in for different types of entities, and can also specify additional patterns that can be used for finding tha...
The Cyc project is predicated on the idea that, in order to be effective and flexible, computer s... more The Cyc project is predicated on the idea that, in order to be effective and flexible, computer software must have an understanding of the context in which its tasks are performed. We believe this context is what is known informally as “common sense.” Over the last twenty years, sufficient common sense knowledge has been entered into Cyc to allow it to more effectively and flexibly support an important task: increasing its own store of world knowledge. In this paper, we describe the Cyc knowledge base and inference system, enumerate the means that it provides for knowledge elicitation, including some means suitable for use by untrained or lightly trained volunteers, review some ways in which we expect to have Cyc assist in verifying and validating collected knowledge, and describe how we expect the knowledge acquisition process to accelerate in the future.
IEEE Transactions on Knowledge and Data Engineering, 1989
The author previously cited an incorrect publisher (see ibid., vol.1, p.84-8, 1989). The correct ... more The author previously cited an incorrect publisher (see ibid., vol.1, p.84-8, 1989). The correct publisher information is: Reading, MA: Addison Wesley
Forty years ago (I'd say "Fifty" but I was still a kid in 1956) we all knew pretty ... more Forty years ago (I'd say "Fifty" but I was still a kid in 1956) we all knew pretty much what 2006 would be like: We'd all have human-level robot servants like Robby (Forbidden Planet), and ubiquitous spoken access to super-human-level disembodied intelligences like the Star Trek Computer ("Working.") and HAL (minus the homicidal mania). The future sure isn't what it used to be! How could so many of us have been so wrong about so much?
■ By extending Cyc’s ontology and knowledge base approximately 2 percent, Cycorp and Cleveland Cl... more ■ By extending Cyc’s ontology and knowledge base approximately 2 percent, Cycorp and Cleveland Clinic Foundation (CCF) have built a system to answer clinical researchers ’ ad hoc queries. The query may be long and complex, hence it is only partially understood at first, parsed into a set of CycL (higher-order logic) fragments with open variables. But, surprisingly often, after applying various constraints (medical domain knowledge, common sense, discourse pragmatics, syntax), there is only one single way to fit those fragments together, one semantically meaningful formal query P. The Semantic Research Assistant (SRA) system dispatches a series of database calls and then combines,
Metaphor is ubiquitous in natural language, and is highly functional, not just decorative. Unfort... more Metaphor is ubiquitous in natural language, and is highly functional, not just decorative. Unfortunately, even the latest encouraging research results in automating its analysis fall well short of the understanding of metaphor required to apply it to cultural analysis. Their coverage is too limited, and they require expensive training data. However, it is a general feature of metaphors that they violate hard or soft semantic constraints. They can be analyzed by detecting that these constraints have been violated, and describing how. Some of the gap in understanding can be closed simply by having a crisp model of argument types for the predicates that are denoted in text – a violation of those types is evidence for a metaphor. Closing the remaining gap, for more ambiguous cases, will require discourse analysis and real world knowledge to eliminate interpretations that are conceivable, but very unlikely. This sort of semantic analysis of metaphor is both possible and practical, and ca...
However, it is a general feature of metaphors that they violate hard or soft semantic constraints... more However, it is a general feature of metaphors that they violate hard or soft semantic constraints. They can be analyzed by detecting that these constraints have been violated, and describing how. Some of the gap in understanding can be closed simply by having a crisp model of argument types for the predicates that are denoted in text – a violation of those types is evidence for a metaphor. Closing the remaining gap, for more ambiguous cases, will require discourse analysis and real world knowledge to eliminate interpretations that are conceivable, but very unlikely. This sort of semantic analysis of metaphor is both possible and practical, and can be done at relatively modest cost by exploiting the state of the art in knowledge bases, inference and scalable shallow text analysis.
Eurisko is an AI program that learns by discovery. We are applying Eurisko to the task of inventi... more Eurisko is an AI program that learns by discovery. We are applying Eurisko to the task of inventing new kinds of three- dimensional microelectronic devices that can then be fabricated using recently developed laser recrystallization techniques. Three experiments have been conducted, and some novel designs and design rules have emerged. The paradigm for Eurisko's exploration is a loop in which it generates a new device configuration, computes its I/O behavior, tries to "parse" this into a functionally it already knows about and can use, and then evaluates the results. In the first experiment, this loop took place at the level of charged carriers moving under the effects of electric fields through abutted regions of doped and undoped semiconductors. Many of the well-known primitive devices were synthesized quickly, such as the MOSFET, Junction Diode, and Bipolar Transistor. This was unsurprising, as they were short sentences in the descriptive language we had defined (a ...
On March 27, 2006, I gave a light-hearted and occasionally bittersweet presentation on “Whatever ... more On March 27, 2006, I gave a light-hearted and occasionally bittersweet presentation on “Whatever Happened to AI?” at the Stanford Spring Symposium presentation – to a lively audience of active AI researchers and formerly-active ones (whose current inaction could be variously ascribed to their having aged, reformed, given up, redefined the problem, etc.) This article is a brief chronicling of that talk, and I entreat the reader to take it in that spirit: a textual snapshot of a discussion with friends and colleagues, rather than a scholarly article. I begin by whining about the Turing Test, but only for a thankfully brief bit, and then get down to my top-10 list of factors that have retarded progress in our field, that have delayed the emergence of a true strong AI.
Abstract : This work was part of the DARPA High Performance Knowledge Base (HPKB) program. The wo... more Abstract : This work was part of the DARPA High Performance Knowledge Base (HPKB) program. The work described in this final report has focused on providing to the HPKB program the robustness and effectiveness of common sense knowledge as embodied in the Cyc knowledge base. Its objective was to provide intermediate level knowledge necessary to tie together high level, abstract knowledge and low level application specific knowledge to ease integration of knowledge bases and provide more efficient and more powerful inferencing mechanisms. The pre-existing Cyc KB had tens of thousands of useful rules for HPKB Integrated Knowledge Base (IKB) to inherit, and the Cyc team had already analyzed the "perennial conceptual issues" for thirteen years prior to HPKB. Early adoption of Cyc's Public Upper Ontology as the "HPKB Jumpstart Ontology" gave both the Cycorp and SAIC teams a uniform, convenient, and reliable environment to add knowledge, ask questions and gather measurements.
The Cyc project is predicated on the idea that effective machine learning depends on having a cor... more The Cyc project is predicated on the idea that effective machine learning depends on having a core of knowledge that provides a context for novel learned information – what is known informally as "common sense." Over the last twenty years, a sufficient core of common sense knowledge has been entered into Cyc to allow it to begin effectively and flexibly supporting its most important task: increasing its own store of world knowledge. In this paper, we present initial work on a method of using a combination of Cyc and the World Wide Web, accessed via Google, to assist in entering knowledge into Cyc. The long-term goal is automating the process of building a consistent, formalized representation of the world in the Cyc knowledge base via machine learning. We present preliminary results of this work and describe how we expect the knowledge acquisition process to become more accurate, faster, and more automated in the future.
The Cyc project is predicated on the idea that effective machine learning depends on having a cor... more The Cyc project is predicated on the idea that effective machine learning depends on having a core of knowledge that provides a context for novel learned information – what is known informally as "common sense." Over the last twenty years, a sufficient core of common sense knowledge has been entered into Cyc to allow it to begin effectively and flexibly supporting its most important task: increasing its own store of world knowledge. In this paper, we present initial work on a method of using a combination of Cyc and the World Wide Web, accessed via Google, to assist in entering knowledge into Cyc. The long-term goal is automating the process of building a consistent, formalized representation of the world in the Cyc knowledge base via machine learning. We present preliminary results of this work and describe how we expect the knowledge acquisition process to become more accurate, faster, and more automated in the future.
This paper presents a novel method, based on the Cyc Knowledge Base and Inference Engine, of gath... more This paper presents a novel method, based on the Cyc Knowledge Base and Inference Engine, of gathering, organizing and sharing information about entities of interest (be they people, organizations , events or some other type of entity). The formal representations used in the Fact Sheets allow users to easily share information with others , run automated queries against the information , and allow the system to attempt to automatically gather and verify information before presenting it to the analyst. The system automatically keeps track of provenance (both which document a fact came from, and who interpreted the document). When gathering information automatically, the system produces a variety of search strings (using all known names for the entity) and then scours its sources for possible answers. Individual analysts can specify what types of information they are interested in for different types of entities, and can also specify additional patterns that can be used for finding tha...
This paper presents a novel method, based on the Cyc Knowledge Base and Inference Engine, of gath... more This paper presents a novel method, based on the Cyc Knowledge Base and Inference Engine, of gathering, organizing and sharing information about entities of interest (be they people, organizations , events or some other type of entity). The formal representations used in the Fact Sheets allow users to easily share information with others , run automated queries against the information , and allow the system to attempt to automatically gather and verify information before presenting it to the analyst. The system automatically keeps track of provenance (both which document a fact came from, and who interpreted the document). When gathering information automatically, the system produces a variety of search strings (using all known names for the entity) and then scours its sources for possible answers. Individual analysts can specify what types of information they are interested in for different types of entities, and can also specify additional patterns that can be used for finding tha...
The Cyc project is predicated on the idea that, in order to be effective and flexible, computer s... more The Cyc project is predicated on the idea that, in order to be effective and flexible, computer software must have an understanding of the context in which its tasks are performed. We believe this context is what is known informally as “common sense.” Over the last twenty years, sufficient common sense knowledge has been entered into Cyc to allow it to more effectively and flexibly support an important task: increasing its own store of world knowledge. In this paper, we describe the Cyc knowledge base and inference system, enumerate the means that it provides for knowledge elicitation, including some means suitable for use by untrained or lightly trained volunteers, review some ways in which we expect to have Cyc assist in verifying and validating collected knowledge, and describe how we expect the knowledge acquisition process to accelerate in the future.
IEEE Transactions on Knowledge and Data Engineering, 1989
The author previously cited an incorrect publisher (see ibid., vol.1, p.84-8, 1989). The correct ... more The author previously cited an incorrect publisher (see ibid., vol.1, p.84-8, 1989). The correct publisher information is: Reading, MA: Addison Wesley
Forty years ago (I'd say "Fifty" but I was still a kid in 1956) we all knew pretty ... more Forty years ago (I'd say "Fifty" but I was still a kid in 1956) we all knew pretty much what 2006 would be like: We'd all have human-level robot servants like Robby (Forbidden Planet), and ubiquitous spoken access to super-human-level disembodied intelligences like the Star Trek Computer ("Working.") and HAL (minus the homicidal mania). The future sure isn't what it used to be! How could so many of us have been so wrong about so much?
■ By extending Cyc’s ontology and knowledge base approximately 2 percent, Cycorp and Cleveland Cl... more ■ By extending Cyc’s ontology and knowledge base approximately 2 percent, Cycorp and Cleveland Clinic Foundation (CCF) have built a system to answer clinical researchers ’ ad hoc queries. The query may be long and complex, hence it is only partially understood at first, parsed into a set of CycL (higher-order logic) fragments with open variables. But, surprisingly often, after applying various constraints (medical domain knowledge, common sense, discourse pragmatics, syntax), there is only one single way to fit those fragments together, one semantically meaningful formal query P. The Semantic Research Assistant (SRA) system dispatches a series of database calls and then combines,
Metaphor is ubiquitous in natural language, and is highly functional, not just decorative. Unfort... more Metaphor is ubiquitous in natural language, and is highly functional, not just decorative. Unfortunately, even the latest encouraging research results in automating its analysis fall well short of the understanding of metaphor required to apply it to cultural analysis. Their coverage is too limited, and they require expensive training data. However, it is a general feature of metaphors that they violate hard or soft semantic constraints. They can be analyzed by detecting that these constraints have been violated, and describing how. Some of the gap in understanding can be closed simply by having a crisp model of argument types for the predicates that are denoted in text – a violation of those types is evidence for a metaphor. Closing the remaining gap, for more ambiguous cases, will require discourse analysis and real world knowledge to eliminate interpretations that are conceivable, but very unlikely. This sort of semantic analysis of metaphor is both possible and practical, and ca...
However, it is a general feature of metaphors that they violate hard or soft semantic constraints... more However, it is a general feature of metaphors that they violate hard or soft semantic constraints. They can be analyzed by detecting that these constraints have been violated, and describing how. Some of the gap in understanding can be closed simply by having a crisp model of argument types for the predicates that are denoted in text – a violation of those types is evidence for a metaphor. Closing the remaining gap, for more ambiguous cases, will require discourse analysis and real world knowledge to eliminate interpretations that are conceivable, but very unlikely. This sort of semantic analysis of metaphor is both possible and practical, and can be done at relatively modest cost by exploiting the state of the art in knowledge bases, inference and scalable shallow text analysis.
Eurisko is an AI program that learns by discovery. We are applying Eurisko to the task of inventi... more Eurisko is an AI program that learns by discovery. We are applying Eurisko to the task of inventing new kinds of three- dimensional microelectronic devices that can then be fabricated using recently developed laser recrystallization techniques. Three experiments have been conducted, and some novel designs and design rules have emerged. The paradigm for Eurisko's exploration is a loop in which it generates a new device configuration, computes its I/O behavior, tries to "parse" this into a functionally it already knows about and can use, and then evaluates the results. In the first experiment, this loop took place at the level of charged carriers moving under the effects of electric fields through abutted regions of doped and undoped semiconductors. Many of the well-known primitive devices were synthesized quickly, such as the MOSFET, Junction Diode, and Bipolar Transistor. This was unsurprising, as they were short sentences in the descriptive language we had defined (a ...
On March 27, 2006, I gave a light-hearted and occasionally bittersweet presentation on “Whatever ... more On March 27, 2006, I gave a light-hearted and occasionally bittersweet presentation on “Whatever Happened to AI?” at the Stanford Spring Symposium presentation – to a lively audience of active AI researchers and formerly-active ones (whose current inaction could be variously ascribed to their having aged, reformed, given up, redefined the problem, etc.) This article is a brief chronicling of that talk, and I entreat the reader to take it in that spirit: a textual snapshot of a discussion with friends and colleagues, rather than a scholarly article. I begin by whining about the Turing Test, but only for a thankfully brief bit, and then get down to my top-10 list of factors that have retarded progress in our field, that have delayed the emergence of a true strong AI.
Abstract : This work was part of the DARPA High Performance Knowledge Base (HPKB) program. The wo... more Abstract : This work was part of the DARPA High Performance Knowledge Base (HPKB) program. The work described in this final report has focused on providing to the HPKB program the robustness and effectiveness of common sense knowledge as embodied in the Cyc knowledge base. Its objective was to provide intermediate level knowledge necessary to tie together high level, abstract knowledge and low level application specific knowledge to ease integration of knowledge bases and provide more efficient and more powerful inferencing mechanisms. The pre-existing Cyc KB had tens of thousands of useful rules for HPKB Integrated Knowledge Base (IKB) to inherit, and the Cyc team had already analyzed the "perennial conceptual issues" for thirteen years prior to HPKB. Early adoption of Cyc's Public Upper Ontology as the "HPKB Jumpstart Ontology" gave both the Cycorp and SAIC teams a uniform, convenient, and reliable environment to add knowledge, ask questions and gather measurements.
Uploads
Papers