Leonardo Lesmo

Evidence combination in expert systems

This paper discusses some of the problems related to the representation of uncertain knowledge and to the combination of evidence degrees in rule-based expert systems. Some of the methods proposed in the literature are briefly analysed with particular attention to the Subjective Bayesian Probability (used in PROSPECTOR) and the Confirmation Theory adopted in MYCIN. The paper presents an integrated approach based on Possibility Theory for evaluating the degree of match between the set of conditions occurring in the antecedent of a production rule and the input data, for combining the evidence degree of a fact with the strength of implication of a rule and for combining evidence degrees coming from different pieces of knowledge. The semantics of the logical operators AND and OR in possibility theory and in our approach are compared. Finally, the definitions of some quantifiers like AT LEAST n, AT MOST n, EXACTLY n are introduced.

A configurable system for the construction of adaptive virtual stores

With the recent expansion of the Internet, the interest towards electronic sales has quickly grown and many tools have been built to help vendors to set up their Web stores. These tools offer all the facilities for building the store databases and managing the order processing and secure payment transactions, but they typically do not focus on issues like the personalization of the interaction with the customers. However, Web surfers are generally heterogeneous and have different needs and preferences; moreover, the trend of marketing strategies is to pay more and more attention to the specific buyers. So, the importance of personalizing the interaction with the user and the product presentation is increasing. In this paper, we describe the architecture of a configurable virtual Web store supporting personalized hypertextual interactions with users. Our system builds a user profile by applying user modeling techniques and stereotypical information about the characteristics of customer groups; this profile is used during the interaction in order to tailor the product descriptions and the selection of items to recommend to the users needs, varying the layout of the hypertextual pages and the detail of the descriptions accordingly. Tailoring the systems behavior requires the parallel execution of several complex tasks during the interaction (e.g., identifying the users preferences, selecting the products most suited to her, dynamically generating the hypertextual pages). Therefore, we have defined a multiagent architecture where these tasks are executed by different agents, which cooperate offering specific services to each other. In our system, the domain‐dependent knowledge, concerning information about products and customer features, is declaratively represented and clearly separated from the domain‐independent components, which represent the core of the virtual store. This separation has the advantage that our architecture can be easily instantiated on several sales domains, therefore obtaining different Web stores out of a single shell. Our system is developed in a Java‐based environment and the overall architecture includes the prototype of a virtual store and the configuration tools which can be used to set up a new store on a specific sales domain.

An Earley-type recognizer for dependency grammar

The paper is a first attempt to fill a gap in the dependency literature, by providing a mathematical result on the complexity of recognition with a dependency grammar. The paper describes an improved Earley-type recognizer with a complexity O(IGI2n3). The improvement is due to a precompilation of the dependency rules into parse tables, that determine the conditions of applicability of two primary actions, predict and scan, used in recognition.

An expert system for the evaluation of liver functional assessment.

The paper describes an expert system for the assessment of the liver function. Since the system must act as an intelligent assistant for a general physician, a major emphasis has been laid upon its interactive capabilities. In particular, the user can ask the system how a given conclusion has been reached (explanation facilities) and can alter the normal operation flow. In the design of the system, the different significance and availability of the clinical data and the laboratory tests have been taken into account: The investigations about a given patient start from the clinical data, and only when there is some evidence of hepatopathy are the results of some laboratory tests requested. The final result of the investigations consists of the assessment of the liver function in terms of four aspects (biosynthesis, cholestasis, cytolysis, reactivity), each of which is assigned a linguistic value describing its impairment degree. The techniques adopted in the system are based on Artifical Intelligence methodologies augmented with linguistic terms handled according to the fuzzy set theory.

Fuzzy Production Rules: A Learning Methodology

In many research fields it is possible to obtain good scientific results only after large amounts of data have been collected and analyzed; the analysis allows the researcher to detect regularities, similarities and discriminant features which may be useful to characterize different classes of objects. On the other hand, the manual examination of a large set of data is slow and error prone, so that many techniques have been proposed and are actually used to perform that analysis automatically (e.g. discriminant analysis); unfortunately, most of those techniques are based on mathematical methodologies which impose strong constraints on the kinds of data that can be analyzed.

Formal Aspects and Parsing Issues of Dependency Theory

The paper investigates the problem of providing a formal device for the dependency approach to syntax, and to link it with a parsing model. After reviewing the basic tenets of the paradigm and the few existing mathematical results, we describe a dependency formalism which is able to deal with long-distance dependencies. Finally, we present an Earley-style parser for the formalism and discuss the (polynomial) complexity results.

Dependency tree semantics

This paper presents Dependency Tree Semantics (DTS), an underspecified logic for representing quantifier scope ambiguities. DTS features a direct interface with a Dependency grammar, an easy management of partial disambiguations and the ability to represent branching quantifier readings. This paper focuses on the syntax of DTS, while does not take into account the model-theoretic interpretation of its well-formed structures.

TULSI: an NLP system for extracting legal modificatory provisions

In this work we present the TULSI system (so named after Turin University Legal Semantic Interpreter), a system to produce automatic annotations of normative documents through the extraction of modificatory provisions. TULSI relies on a deep syntactic analysis and a shallow semantic interpreter that are illustrated in detail. We report the results of an experimental evaluation of the system and discuss them, also suggesting future directions for further improvement.

LEARNING OF FUZZY PRODUCTION RULES FOR MEDICAL DIAGNOSIS

The methods used so far to develop systems for medical consultation relied either on classical probabilistic and pattern recognition schemata or on techniques developed in the field of Artificial Intelligence. The statistical approach guarantees the availability of efficient learning algorithms, but the structure of the decision rules is too far from the methods used by the physicians, thus preventing the designer from inserting good explanation facilities into the system. On the other hand, artificial intelligence systems generally require a noticeable effort to encode the experts knowledge in a form suitable to perform inferences on the available data. The system described in this paper tries to combine the respective advantages of the mentioned approaches, by using fuzzy production rules as the basic deductive mechanism. The paper outlines an algorithm for the automatic learning of the fuzzy production rules and shows its application to the domain of liver pathology. The effectiveness of the learning algorithm is shown by reporting a set of experimental results which allow to compare the outcome of the learned production rules with the classification available from the experts.

On the Ontological Status of Plans and Norms

This article describes an ontological model of norms. The basic assumption is that a substantial part of a legal system is grounded on the concept of agency. Since a legal system aims at regulating a society, then its goal can be achieved only by affecting the behaviour of the members of the society. We assume that a society is made up of agents (which can be individuals, institutions, software programs, etc.), that agents have beliefs, goals and preferences, and that they commit to intentions in order to choose a line of behaviour. The role of norms, within a legal system, is to specify how and when the chosen behaviour agrees with the basic principles of the legal system. In this article, we show how a model based on plans can be the basis for the ontological representation of norms, which are expressed as constraints on the possible plans an agent may choose to guide its behaviour. Moreover, the paper describes how the proposed model can be linked to the upper level of a philosophically well-founded ontology (DOLCE); in this way, the model is set in a wider perspective, which opens the way to further developments.