Pronunciation by analogy (PbA) is a data-driven method for relating letters to sound, with potential application to next-generation text-to-speech systems. This paper extends previous work on PbA in several directions. First, we have included `full' pattern matching between input letter string and dictionary entries, as well as including lexical stress in letter-to-phoneme conversion. Second, we have extended the method to phonemeto -letter conversion. Third, and most important, we have experimented with multiple, different strategies for scoring the candidate pronunciations. Individual scores for each strategy are obtained on the basis of rank and either multiplied or summed to produce a final, overall score. Five strategies have been studied and results obtained from all 31 possible combinations. The two combination methods perform comparably, with the product rule only very marginally superior to the sum rule. Nonparametric statistical analysis reveals that performance improves as more strategies are included in the combination: this trend is very highly significant ( p 0 0005). Accordingly for letter-to-phoneme conversion, best results are obtained when all five strategies are combined: word accuracy is raised to 65.5% relative to 61.7% for our best previous result and 63.0% for the best-performing single strategy. These improvements are very highly significant ( p 0 and p 0 00011 respectively). Similar results were found for phoneme-to-letter and letter-to-stress conversion, although the former was an easier problem for PbA than letter-to-phoneme conversion and the latter was harder. The main sources of error for the multi-strategy approach are very similar to those for the best single strategy, and mostly involve vowel letters and phonemes. 1

The automatic derivation of word pronunciations from input text is a central task for any text-to-speech system. For general English text at least, this is often thought to be a solved problem, with manually-derived linguistic rules assumed capable of handling `novel' words missing from the system dictionary. Data-driven methods, based on machine learning of the regularities implicit in a large pronouncing dictionary, have received considerable attention recently but are generally thought to perform less well. However, these tentative beliefs are at best uncertain without powerful methods for comparing text-to-phoneme subsystems. This paper contributes to the development of such methods by comparing the performance of four representative approaches to automatic phonemisation on the same test dictionary. As well as rule-based approaches, three data-driven techniques are evaluated: pronunciation by analogy (PbA), NETspeak and IB1-IG (a modified k-nearest neighbour method). Issues involved in comparative evaluation are detailed and elucidated. The data-driven techniques outperform rules in accuracy of letter-to-phoneme translation by a very significant margin but require aligned text-phoneme training data and are slower. Best translation results are obtained with PbA at approximately 72% words correct on a reasonably large pronouncing dictionary, compared to something like 26% words correct for the rules, indicating that automatic pronunciation of text is not a solved problem.

A Process Representation Technique proposed in this article can be used for evaluating software processes. The Technique considers some important aspects of causal dependencies between activities and documents within a process model. Production rules are used as the description language, because they are able to handle incomplete and changeable information and to be translated into Prolog predicates in order to execute the evaluation. Key words: process modelling, process models, software evaluation, standards, system views, System refinements, Petri Net notion, rule-based system, conformance testing, Yourdon Structured Method. 1 Introduction The most important aim of software engineering discipline is to improve the quality of a software product. In order to achieve this goal quality requirements have to be considered and evaluated. The evaluation steps can be performed after the development of the software or during the development process. Deviations from the quality requirements ...

For specifying and designing an evaluation and certification of software a number of different types of problems (represented as sub-models) have to be identified: - software characteristics and metrics, - software product information, - software process information, - evaluation methods and tools. The collection of information required for an evaluation and certification is described by an evaluation scenario and comprises part of the evaluation process. A formalised description of an evaluation and certification, called a Formal Model, has been developed which consists of the four sub-models and an Evaluation Advisory Concept. This brings the information together for the evaluation process. The Formal Model reflects the capabilities of an evaluation procedure, as presented in "The Evaluator's Guide" currently being discussed within ISO/IEC JTC1/WG 6 for consideration as an international standard. Keywords software (product) evaluation, evaluation module, evaluation level, quality ch...

Once talking about the possibilities of creating `thinking machines', Turing properly said that this task should start by defining of what is understood as `machines' and of what is `think'. Regarding the pursuit of modeling intelligence, two large avenues were opened by researchers in almost the same epoch: Symbolic Artificial Intelligence -- SAI -- and Connectionist Artificial Intelligence -- CAI -- based respectively on symbols and rules and in artificial neurons. It seems that the time is come to start thinking (and acting) to establish a standard of comparison, that could objectively tell how far we have gone along the road of constructing ever better AI systems. Devising an Intelligence Quotient IQ -- for machines or any intelligent system would be, perhaps, an advancement but unfortunately, the history of the development of techniques to measure human IQ, the first source checked to find applications to AI, points to a very fuzzy zone. Admitting that possibility, we present some conjectures. For example, introducing some metric to evaluate the redundancy of the rules of an Expert System, or the efficiency of a given topology in a Neural Net could bring new insights on ranking AI paradigms and indicate which are the most promising ones.

Many real world conflicts can be understood in terms of logical inconsistencies between goals. In this paper, I present an approach to conflict resolution that unifies logic, goalreduction and condition-action rules in a cognitive model of intelligent agent. The approach uses goal hierarchies to reconcile goal conflicts by finding alternative, logically consistent ways of solving higher-level goals. It also incorporates the use of decision theory to decide between different solutions, in the attempt to optimise their expected utility. To illustrate and test the power and generality of the approach, I investigate two applications: the prisoner's dilemma and the Agha-Malley proposed solution of the Israeli-Palestinian conflict.

Proceedings

Abstract. Electronic bank transactions are very common today. Services given by an Automatic Teller Machine (ATM), for example, are very popular and widely used by bank clients. Unfortunately, in the same way as the use of these devices is increasing, the proliferation of different frauds to try to violate these systems to steal user’s money is also increasing. Sometimes, the modus operandi used by the delinquents depends on different factors, such as the country or the city where fraud is committed or, as in the case of ATMs, the model or location of these devices. Since the detection of these modus operandi is not easy and they could be different from a bank institution to another, having both an environment capable of following up the swindler agents learning processes and a way to prevent the cooperation between these agents to share the learned knowledge, would be very useful to discover different modus operandi before crimes are committed. In this paper, a framework designed to follow up the swindlers ’ agents learning process and to share the knowledge between the agents is presented. This framework is based on the FIPA (Foundation for Intelligent Physical Agents) specifications and it emphasizes on the swindler agents learning process to fulfil the human-like agent behaviour and a realistic interaction with the environment. 1