Specification-based software development makes software easier to validate and maintain. Yet spec... more Specification-based software development makes software easier to validate and maintain. Yet specifications of large systems are themselves large, making understanding and validation difficult. One cause for this problem is that specifications and requirements are kept distinct. This paper describes an approach to specification development in which the specification arises naturally through the requirements analysis process. The emerging specification is developed
Adaptive Agents and Multi-Agents Systems, Jul 15, 2002
Agents (adaptive or intelligent agents and multi-agent systems) are one of the most prominent and... more Agents (adaptive or intelligent agents and multi-agent systems) are one of the most prominent and attractive technologies in Computer Science at the beginning of the new century. Agent and MAS technologies, methods, and theories are currently contributing to many diverse domains such as information retrieval, user interface design, robotics, electronic commerce, computer mediated collaboration, computer games, education and training, smart environments, ubiquitous computers, social simulation, etc.Agents are not only a very promising technology, they are emerging as a new way of thinking, a conceptual paradigm for analyzing problems and for designing systems, for dealing with complexity, distribution and interactivity, and perhaps a new perspective on computing and intelligence. Yet to realize this promise further advances are required in agent architectures, languages, theories, and design techniques.To accept this challenge, three very successful and high level events, AA (the International Conference on Autonomous Agents), ICMAS (the International Conference on Multi-Agent Systems), and ATAL (the International Workshop on Agent Theories, Architectures, and Languages), decided to build on their past successes and converge in 2002 in a Joint Conference.The aims of this joint venture are: to strengthen the relationships within the broader Agent community, to enhance quality while promoting innovation, to encourage richer exchanges between theoretical and experimental research and applications, modeling and engineering, and between the micro and the macro levels of agent design, and to foster collaboration between information science and technology and the cognitive and social sciences. The theme of the conference is Bringing People and Agents Together. As agent-based computing becomes ever more widespread, it is increasingly important for agents to interact effectively with people and with other agents. In addition to the normal research topics of the component conferences, we particularly encouraged submissions that address issues of how to improve interaction and collaboration among agents and people.Response to the new conference was very favorable. The over 500 papers were submitted to the conference. Some of these were selected for publication as full papers, and others were selected for presentation as posters. The acceptance rate for full papers was 27%, low enough to ensure high quality, yet high enough to include a variety of topics and perspectives. The paper and poster summaries are included in this proceedings, which is divided into three volumes, one for each day of the conference.
Adaptive Agents and Multi-Agents Systems, Jul 15, 2002
Agents (adaptive or intelligent agents and multi-agent systems) are one of the most prominent and... more Agents (adaptive or intelligent agents and multi-agent systems) are one of the most prominent and attractive technologies in Computer Science at the beginning of the new century. Agent and MAS technologies, methods, and theories are currently contributing to many diverse domains such as information retrieval, user interface design, robotics, electronic commerce, computer mediated collaboration, computer games, education and training, smart environments, ubiquitous computers, social simulation, etc.Agents are not only a very promising technology, they are emerging as a new way of thinking, a conceptual paradigm for analyzing problems and for designing systems, for dealing with complexity, distribution and interactivity, and perhaps a new perspective on computing and intelligence. Yet to realize this promise further advances are required in agent architectures, languages, theories, and design techniques.To accept this challenge, three very successful and high level events, AA (the International Conference on Autonomous Agents), ICMAS (the International Conference on Multi-Agent Systems), and ATAL (the International Workshop on Agent Theories, Architectures, and Languages), decided to build on their past successes and converge in 2002 in a Joint Conference.The aims of this joint venture are: to strengthen the relationships within the broader Agent community, to enhance quality while promoting innovation, to encourage richer exchanges between theoretical and experimental research and applications, modeling and engineering, and between the micro and the macro levels of agent design, and to foster collaboration between information science and technology and the cognitive and social sciences. The theme of the conference is Bringing People and Agents Together. As agent-based computing becomes ever more widespread, it is increasingly important for agents to interact effectively with people and with other agents. In addition to the normal research topics of the component conferences, we particularly encouraged submissions that address issues of how to improve interaction and collaboration among agents and people.Response to the new conference was very favorable. The over 500 papers were submitted to the conference. Some of these were selected for publication as full papers, and others were selected for presentation as posters. The acceptance rate for full papers was 27%, low enough to ensure high quality, yet high enough to include a variety of topics and perspectives. The paper and poster summaries are included in this proceedings, which is divided into three volumes, one for each day of the conference.
We are exploring the use of virtual reality for training people how to perform tasks, such as ope... more We are exploring the use of virtual reality for training people how to perform tasks, such as operating and maintaining complex equipment. This video describes Steve, an agent we are developing that assists in the training. Steve is an autonomous, animated agent that cohabits the virtual world with students. Steve continuously monitors the state of the virtual world, periodically manipulating it through virtual motor actions. His objective is to help students learn to perform physical, procedural tasks. He can demonstrate tasks, explaining his actions, as well as monitor students performing tasks, providing help when they need it. In addition to teaching students individual tasks, he can also help them learn to perform multi-person team tasks: he can serve as a tutor for a student learning a particular role in the team, and he can play the role of a teammate when a human teammate is unavailable. By integrating previous work in agent architectures, intelligent tutoring systems, and computer graphics, Steve illustrates a new breed of computer tutor: a human-like agent that can interact with students in a virtual world to help them learn.
Specification-based software development makes software easier to validate and maintain. Yet spec... more Specification-based software development makes software easier to validate and maintain. Yet specifications of large systems are themselves large, making understanding and validation difficult. One cause for this problem is that specifications and requirements are kept distinct. This paper describes an approach to specification development in which the specification arises naturally through the requirements analysis process. The emerging specification is developed
Adaptive Agents and Multi-Agents Systems, Jul 15, 2002
Agents (adaptive or intelligent agents and multi-agent systems) are one of the most prominent and... more Agents (adaptive or intelligent agents and multi-agent systems) are one of the most prominent and attractive technologies in Computer Science at the beginning of the new century. Agent and MAS technologies, methods, and theories are currently contributing to many diverse domains such as information retrieval, user interface design, robotics, electronic commerce, computer mediated collaboration, computer games, education and training, smart environments, ubiquitous computers, social simulation, etc.Agents are not only a very promising technology, they are emerging as a new way of thinking, a conceptual paradigm for analyzing problems and for designing systems, for dealing with complexity, distribution and interactivity, and perhaps a new perspective on computing and intelligence. Yet to realize this promise further advances are required in agent architectures, languages, theories, and design techniques.To accept this challenge, three very successful and high level events, AA (the International Conference on Autonomous Agents), ICMAS (the International Conference on Multi-Agent Systems), and ATAL (the International Workshop on Agent Theories, Architectures, and Languages), decided to build on their past successes and converge in 2002 in a Joint Conference.The aims of this joint venture are: to strengthen the relationships within the broader Agent community, to enhance quality while promoting innovation, to encourage richer exchanges between theoretical and experimental research and applications, modeling and engineering, and between the micro and the macro levels of agent design, and to foster collaboration between information science and technology and the cognitive and social sciences. The theme of the conference is Bringing People and Agents Together. As agent-based computing becomes ever more widespread, it is increasingly important for agents to interact effectively with people and with other agents. In addition to the normal research topics of the component conferences, we particularly encouraged submissions that address issues of how to improve interaction and collaboration among agents and people.Response to the new conference was very favorable. The over 500 papers were submitted to the conference. Some of these were selected for publication as full papers, and others were selected for presentation as posters. The acceptance rate for full papers was 27%, low enough to ensure high quality, yet high enough to include a variety of topics and perspectives. The paper and poster summaries are included in this proceedings, which is divided into three volumes, one for each day of the conference.
Adaptive Agents and Multi-Agents Systems, Jul 15, 2002
Agents (adaptive or intelligent agents and multi-agent systems) are one of the most prominent and... more Agents (adaptive or intelligent agents and multi-agent systems) are one of the most prominent and attractive technologies in Computer Science at the beginning of the new century. Agent and MAS technologies, methods, and theories are currently contributing to many diverse domains such as information retrieval, user interface design, robotics, electronic commerce, computer mediated collaboration, computer games, education and training, smart environments, ubiquitous computers, social simulation, etc.Agents are not only a very promising technology, they are emerging as a new way of thinking, a conceptual paradigm for analyzing problems and for designing systems, for dealing with complexity, distribution and interactivity, and perhaps a new perspective on computing and intelligence. Yet to realize this promise further advances are required in agent architectures, languages, theories, and design techniques.To accept this challenge, three very successful and high level events, AA (the International Conference on Autonomous Agents), ICMAS (the International Conference on Multi-Agent Systems), and ATAL (the International Workshop on Agent Theories, Architectures, and Languages), decided to build on their past successes and converge in 2002 in a Joint Conference.The aims of this joint venture are: to strengthen the relationships within the broader Agent community, to enhance quality while promoting innovation, to encourage richer exchanges between theoretical and experimental research and applications, modeling and engineering, and between the micro and the macro levels of agent design, and to foster collaboration between information science and technology and the cognitive and social sciences. The theme of the conference is Bringing People and Agents Together. As agent-based computing becomes ever more widespread, it is increasingly important for agents to interact effectively with people and with other agents. In addition to the normal research topics of the component conferences, we particularly encouraged submissions that address issues of how to improve interaction and collaboration among agents and people.Response to the new conference was very favorable. The over 500 papers were submitted to the conference. Some of these were selected for publication as full papers, and others were selected for presentation as posters. The acceptance rate for full papers was 27%, low enough to ensure high quality, yet high enough to include a variety of topics and perspectives. The paper and poster summaries are included in this proceedings, which is divided into three volumes, one for each day of the conference.
We are exploring the use of virtual reality for training people how to perform tasks, such as ope... more We are exploring the use of virtual reality for training people how to perform tasks, such as operating and maintaining complex equipment. This video describes Steve, an agent we are developing that assists in the training. Steve is an autonomous, animated agent that cohabits the virtual world with students. Steve continuously monitors the state of the virtual world, periodically manipulating it through virtual motor actions. His objective is to help students learn to perform physical, procedural tasks. He can demonstrate tasks, explaining his actions, as well as monitor students performing tasks, providing help when they need it. In addition to teaching students individual tasks, he can also help them learn to perform multi-person team tasks: he can serve as a tutor for a student learning a particular role in the team, and he can play the role of a teammate when a human teammate is unavailable. By integrating previous work in agent architectures, intelligent tutoring systems, and computer graphics, Steve illustrates a new breed of computer tutor: a human-like agent that can interact with students in a virtual world to help them learn.
Uploads
Papers by Lewis Johnson