A genetic algorithm is used to find the keys of Type II periodic polyalphabetic ciphers with mixed primary alphabets. Because of the difficulty of the ciphertext only cryptanalysis for Type II ciphers, a parallel, multi-phased search strategy is used, each phase of which recovers a bigger portion of the key. See also Trinity College Cryptology Website

R. A. Morelli, R.E.Walde, and W. Servos (2004). A study of heuristic search algorithms for breaking short cryptograms. International Journal of Artificial Intelligence Tools (IJAIT), 13(1), pp. 45-64.

In this study, we compare the use of genetic algorithms (GAs) and other forms of heuristic search in the cryptanalysis of short cryptograms. This paper expands on the work presented at FLAIRS-2003, which established the feasibility of a word-based genetic algorithm (GA) for analyzing short cryptograms. In this study the following search heuristics are compared both theoretically and experimentally: hill-climbing, simulated annealing, word-based and frequency-based genetic algorithms. Although the results reported apply to substitution ciphers in general, we focus in particular on short substitution cryptograms, such as the kind found in newspapers and puzzle books. Short cryptograms present a more challenging form of the problem. The word-based approach uses a relatively small dictionary of frequent words. The frequency-based approaches use frequency data for 2-, 3- and 4-letter sequences. The study shows that all of the optimization algorithms are successful at breaking short cryptograms, but perhaps more significantly, the most important factor in their success appears to be the choice of fitness measure employed. See also Trinity College Cryptology Website

R.A. Morelli and R.E. Walde (2003). A word-based genetic algorithm for cryptanalysis of short cryptograms. Proceedings of the 2003 Florida Artificial Intelligence Research Symposium (FLAIRS-2003). pp. 229-233.

This paper demonstrates the feasibility of a word-based genetic algorithm (GA) for solving short substitution cryptograms such as the kind found in newspapers and puzzle books. Although GA's based on analysis of letter, digram, or trigram frequencies have been used on substitution cryptograms, they are not able to solve short (10-30 word) cryptograms of the sort we address. By using a relatively small dictionary of frequent words to initialize a set of substitution keys, and by employing a word-based crossing mechanism, the GA achieves performance that is comparable to deterministic word-based algorithms. See also Trinity College Cryptology Website

R. Morelli. (1997), Using Machine Learning to Acquire Domain Knowledge in an Intelligent Tutoring System. In Proceedings of the Tenth Florida AI Research Symposium, pp. 26-30, April, 1997.

This paper describes a prototype system that uses machine learning to acquire domain knowledge for an Intelligent Tutoring System in propositional logic. The system uses a form of inductive learning to learn a hierarchy of operators by taking instructions from a human expert. It successfully demonstrates that domain knowledge can be effectively acquired through instruction and integrated into a tutorial framework.

R. Morelli. (1997), Teaching the tutor (extended abstract). In Proceedings Educational Multimedia/Hypermedia and Telecommunications, Vol. II, pp. 1324-1325, May 1997. Read in html format.

R. Morelli. (1997), A WWW Client/Server framework for educational software (extended abstract). In Proceedings Educational Multimedia/Hypermedia and Telecommunications, Vol. II, pp. 1326-1327, May 1997. Read in html format.

R. Morelli, B. Dinkins and G. Pelton*. (1995). A tutoring architecture that learns (abstract). In Proceedings of AIED-95: World Conference on Artificial Intelligence in Education, Washington, DC, p. 186.

*School of Computer Science
Carnegie Mellon University

Soar/ITS is an architecture for Intelligent Tutoring Systems (ITSs) based on Soar. As an experiment in ITS design, the main question addressed by Soar/ITS is whether machine learning can be used effectively to develop a practical tutoring architecture. We want a tutoring architecture that is more flexible than current tutoring systems both in terms of how it presents lessons and in how it responds to what the student is doing. This flexibility, however, should not come at the expense of efficiency or tractability. To meet these goals we the architecture must achieve substantial transfer of what it learns across different tasks. We show that learning in Soar/ITS results in considerable transfer and speedup across routine tasks. This gives us reason to believe that the architecture will be practical. More importantly, we show that Soar/ITS is able to transfer learning from routine problem solving tasks into the tutoring task. In particular, Soar/ITS is able to transfer knowledge learned through solving domain problems -- e.g, solving a problem in electrostatics -- into the student monitoring task -- e.g., observing the student solve a similar problem. We believe this type of transfer will allow us to approach important ITS tasks, such as student modeling, with a set of domain independent tools. After describing the Soar/ITS's main architectural features, we provide three detailed examples of effective use of its learning mechanism and show how these can be developed in ways that significantly enhance the performance of student modeling and tutor control. Read in html format.

Acknowledgement: This work was supported in part by a Research Opportunity Award from the National Science Foundation (Subgrant #513566-55455 of NSF grant MDR-9150008).

R. Morelli. (1993), Soar/ITS: A perception-driven intelligent tutoring system architecture, In Proceedings of the Sixth Florida AI Research Symposium, pp. 315-319, April, 1993.

This paper describes Soar/ITS, a prototype intelligent tutoring architecture based on Soar. Soar is a computational paradigm capable of modeling human cognition at a variety of levels. Three novel features of Soar/ITS are described: (1) its ability to represent task knowledge in a flexible and open-ended manner; (2) its ability to represent perceptual and motor knowledge in a cognitively plausible way; and (3), its ability to learn new representations of both the task and the student during the tutoring process. It is shown how these features address some of the outstanding problems in current ITS research. Read in html format.

R. A. Morelli, J.D. Bronzino, J.W.Goethe*. (1993), Conversations for Action: A Speech Act Model of Human-Computer Communication in a Psychiatric Hospital , Journal of Intelligent Systems, Vol. 3, No. 2-4, pp. 87-118.

* The Institute of Living
Hartford, CT 06106

When a staff physician says to an intern he is supervising ÒI think you should try medication X,Ó this utterance may differ in meaning from the same string of words spoken to another staff physician. In the first case, the statement may have the force of an order (ÒDo this!Ó), while in the latter it is merely a suggestion. In either case, the utterance sets up important expectations which constrain the future actions of the parties involved. This paper presents an analytic framework, based on speech act theory, for representing such ÒconversationsÓ so that they may be used in the design of a computer system. Our design perspective views the information system -- in this case an expert system that monitors drug treatment -- as one of many "agents" within a broad communicative network. Speech act theory is used to model the communicative actions within a psychiatric hospital unit and portions of the resulting model are used to support various design decisions. We have found this approach to be useful for addressing a variety of design issues ranging from the specification of input/output screens to the clarification of the systemÕs potential impact on the physicianÕs decision making role and responsibility.

R. A. Morelli. (1990). The Student as Knowledge Engineer: A Constructivist Model for Science Education. Journal of Computing In Higher Education, 2(1), 78-102.

Knowledge engineering is the process whereby a knowledge engineer works closely with a domain expert to build an expert system. The process itself is known to improve both the knowledge engineer's and the expert's understanding of the domain. Moreover, building an expert system is a lot like constructing a scientific theory: both activities result in the creation of an explanatory or problem-solving model of some particular domain. Mindful of this, an attempt was made to exploit the pedagogical potential of the knowledge engineering process by using it as a means of "teaching" a group of junior high school students how to do botanical classification. Serving as knowledge engineers, the students developed an expert advisory system capable of identifying tree specimens from descriptions of their gross morphology. Our evaluations indicated that the students not only mastered the target knowledge, but also enjoyed the opportunity to take a somewhat different approach to this standard junior high school subject. The success of this experiment supports the claim that the knowledge engineering process can serve as an innovative model for science education at the secondary or undergraduate levels. The model encourages a more creative, constructivist approach to teaching certain science concepts and skills, while at the same time fostering the improvement of logic, communication and independent learning skills.

R. A. Morelli. (1990). Using knowledge engineering to teach science. IEEE Expert, 5(4), 74-78.

Knowledge engineering, during which knowledge engineers and domain experts struggle to build precise representations of target domains, can serve as an innovative model for teaching science. The process improves the domain understanding of knowledge engineers and domain experts. Mind of this, we attempted to exploit knowledge engineeringÕs pedagogical potential by engaging high school students in an expert system project at Trinity College. Serving as knowledge engineers, the students developed an expert system for identifying tree species. Project evaluation indicates that students mastered target knowledge. Also, they enjoyed taking a somewhat different approach to this standard junior high school subject. The success of the experiment supports our view that knowledge engineering should be considered as an innovative approach for teaching scientific subjects and improving independent learning skills.