Prof. Gheorghe Tecuci

Dr. Gheorghe Tecuci

Professor of Computer Science and Director of the Learning Agents Center
MSN 6B3, Learning Agents Center, Long and Kimmy Nguyen Engineering Building, Room 4613
The Volgenau School of Engineering
George Mason University, Fairfax, VA 22030
email: tecuci at gmu dot edu, tel: (703) 993-1722,

Teaching and Research Interests: Artificial Intelligence, Knowledge Engineering, Machine Learning, Knowledge Acquisition, Evidence-Based Reasoning, Intelligence Analysis, Semantic Web, Cognitive Assistants, Expert Systems, Intelligent Tutoring Systems

Education: Gheorghe Tecuci received his M.S. degree in Computer Science from the Polytechnic Institute of Bucharest in 1979, graduating first among all the Computer Science students at the Polytechnic Universities of Romania. He received two Ph.D. degrees in Computer Science, one from the University of Paris-South, in July 1988, and the other one from the Polytechnic Institute of Bucharest, in December 1988.

Positions: Between 1979 and 1989 Dr. Tecuci was a researcher at the Romanian Institute for Informatics. He joined the faculty of George Mason University in 1990 as an Assistant Professor, became Associate Professor in 1993, and Professor in 1996. Between 1994 and 1999, Dr. Tecuci was also Coordinating director of the Center for Machine Learning, Natural Language Processing and Conceptual Modeling of the Romanian Academy. In 1995 he established the Learning Agent Laboratory at George Mason University which became Learning Agents Center in 2004. He was Chair of Artificial Intelligence at the US Army War College between 2001 and 2003, and visiting professor between 2001 and 2011.

Research Contributions: He has published around 200 papers (, contributing to new research directions in artificial intelligence, including: multistrategy learning (Disciple: A Theory, Methodology and System for Learning Expert Knowledge, Université de Paris-Sud, 1988; Special Issue of Informatica on Multistrategy Learning, The Slovene Society Informatika, 1993; Proceedings of the First and of the Second International Workshops on Multistrategy Learning, GMU Center for Artificial Intelligence, 1991, 1993; Multistrategy Learning, Morgan Kaufmann, 1994, the last three in collaboration with Ryszard Michalski), integration of machine learning and knowledge acquisition (Special Issue of the Knowledge Acquisition Journal on the Integration of Machine Learning and Knowledge Acquisition, Academic Press, 1994, in colab.; Machine Learning and Knowledge Acquisition: Integrated Approaches, Academic Press, 1995, in colaboration with Yves Kodratoff), instructable agents (Building Intelligent Agents: An Apprenticeship Multistrategy Learning Theory, Methodology, Tool and Case Studies, Academic Press, 1998), mixed-initiative reasoning (Special Issue of the AI Magazine on Mixed-Initiative Assistants, AAAI Press, 2007, in colaboration with Mihai Boicu and Michael Cox), center of gravity analysis (Agent-Assisted Center of Gravity Analysis, George Mason University Press, 2008, in colaboration with Mihai Boicu and Jerome Comello), knowledge engineering and cognitive assistants (Knowledge Engineering: Building Cognitive Assistants for Evidence-based Reasoning, Cambridge University Press, 2016, in colaboration with Dorin Marcu, Mihai Boicu and David Schum), and computational theory of evidence-based reasoning (Intelligence Analysis as Discovery of Evidence, Hypotheses, and Arguments: Connecting the Dots, Cambridge University Press, 2016, in colaboration with David Schum, Dorin Marcu, and Mihai Boicu).

He has followed a career-long interest in the creation of a computational theory and technology for the development of cognitive assistants that enable a seamless synergistic integration of the complementary reasoning abilities of humans and computers. These software agents can learn complex problem-solving expertise directly from human experts, can support experts and non-experts in problem solving and decision making, and can teach their problem-solving expertise to students. Together with his students and collaborators, especially Mihai Boicu and Dorin Marcu, he has continuously developed this approach to agent development, called the Disciple approach. More recently, with the additional contribution of David Schum, the Disciple approach he has extended with a computational theory of evidence-based reasoning. This agent theory and technology contributes to a new revolution in computer science by enabling typical computer users to develop cognitive assistants that incorporate their expertise and help them cope with the challenges of an increasingly complex knowledge society.

Agent Development and Applications: With Mihai Boicu and Dorin Marcu, as well as his students, Dr. Tecuci has successfully employed the Disciple approach to develop and apply knowledge-based agents for a wide variety of domains. Disciple-EBR is a general learning agent shell for evidence-based reasoning which consists of a suite of software tools for the development of specialized Disciple knowledge-based intelligent agents. It includes multiple modules for problem solving, learning, tutoring, evidence-based reasoning, mixed-initiative interaction, as well as a hierarchically organized knowledge base with domain-independent knowledge for evidence-based reasoning at the top of the knowledge hierarchy. A version of this shell is distributed with the book by Tecuci G., Marcu D., Boicu M., Schum D.A., Knowledge Engineering: Building Cognitive Assistants for Evidence-based Reasoning, Cambridge University Press, 2016. Disciple-COG is a learning and decision-support agent for military center of gravity determination. This software system has been used by high ranking military officers in courses at the US Army War College (since 2001), at the Air War College and at George Mason University. The most recent version of it is available on CD as part of the book by Tecuci G., Boicu M., and Comello J., Agent-Assisted Center of Gravity Analysis, GMU Press, 2008. Disciple-LTA is a new type of tool that incorporates a significant amount of knowledge from the Science of Evidence. It can rapidly acquire expertise in intelligence analysis, can train new intelligence analysts, and can help analysts find solutions to complex problems, through mixed-initiative reasoning, allowing a synergistic integration of a human's experience and creativity with an agent's knowledge and speed. Successive versions of this system have been used in courses at the US Army War College and at George Mason University, as well as in experiments and special courses with intelligence analysts. Disciple-FS is a suite of tools for building, utilizing, and maintaining regulatory knowledge bases for modern, dynamic business organizations, especially financial services firms. It has been developed as part of an NSF-funded Small Business Technology Transfer Program and has been licensed to Exprentis Inc. TIACRITIS is a web agent for teaching intelligence analysts the critical thinking skills needed to perform evidence-based reasoning. This system (and associated textbook) is based on a computational theory which views Intelligence Analysis as ceaseless discovery of evidence, hypotheses, and arguments, in a complex world that is changing all the time. It helps students learn about the properties, uses, and marshaling of evidence upon which all analyses rest, through regular practice involving analyses of evidence in both hypothetical and real situations. Disciple-CD (cognitive assistant for connecting the dots) is a knowledge based system for evidence-based hypotheses analysis. It is used in several courses and by several government organizations. It is also the system distributed with the book by Tecuci G., Schum D.A., Marcu D., Boicu M., Intelligence Analysis as Discovery of Evidence, Hypotheses, and Arguments: Connecting the Dots, Cambridge University Press, 2016. Disciple-CD was licensed to Uncharted Inc. Cogent (Cognitive Agent for Cogent Analysis) is a new generation tool for evidence-based hypothesis analysis. As compared to Disciple-CD, it supports a more general argumentation structure, with both favoring and disfavoring arguments for each hypothesis and sub-hypothesis. It has a drastically simplified interface and interaction to facilitate it use by less technical users. It combines first order logic with Baconian and Fuzzy probabilities. Other developed systems are Disciple-WA (for military engineering planning), Disciple-COA (for course of action critiquing), and TIACRITIS-VE (for modeling the behavior of violent extremists).

Support: Since joining George Mason University, G. Tecuci has been awarded over $10 million as Principal Investigator / Program Director, and over $2 million as Co-Principal Investigator. The awards were made by prestigious funding organizations such as the Defense Advanced Research Projects Agency, the National Geospatial-Intelligence Agency, the Air Force Office of Scientific Research, the Air Force Research Lab, the National Science Foundation, the Department of Defense, the US Army, and others.

Recognitions: Dr. Tecuci was elected member of the Romanian Academy and has received several awards, including the US Army Outstanding Civilian Service Medal ("for groundbreaking contributions to the application of Artificial Intelligence to Center of Gravity determination"), the IT&E Outstanding Research Faculty Award, the Best Paper Award at the International Conference on Intelligent Tutoring Systems, and the Deployed Innovative Application Award from the American Association for Artificial Intelligence. Dr. Tecuci has been a guest editor or on the editorial board of 12 journals, and acted as chairman, organizer, or program committee member of over 70 conferences and workshops.

Learning Agents Center
Computer Science Department