Abstract
This paper presents a method for modelling and monitoring the individual decision process of wine tasting panelllists. The individual decision of approving or rejecting a wine is based on the panellist's evaluation of the wine characteristics. The decision process is modelled through a logistic regression where the dependent variable is the decision itself (rejected-0 or approved-1) and the independent variables are the characteristics that were considered relevant for the wine evaluation. The process is monitored using a CUSUM control chart for binomial proportions. The control chart variable is the number of disagreements observed (in a sample of size N) between the actual decisions of the panellist and the results anticipated by the logistic regression model. The method was tested in the Port Wine Institute. Such a test revealed that the logistic model is a useful tool for clarifying both the way in which the wine evaluation process is carried out, and how the decisions are taken. Furthermore, the test has also showed that the control chart is responsive to significant changes of the individual decision process.

Abstract
The Intensive Care Unit (ICU) of Hospitals deals with patients in life critical conditions. The Intensive Care Information System (ICS) can therefore provide extremely important information to support medical doctors' (MDs) decisions. For instance, it is critical to manage well information about the evolution of a large amount of infections over time, about the antibiotics administered to each patient, and the impact on his/her life condition. Good quality information and interaction in such an extreme environment is therefore critical for helping MDs target well medicines to patients. This paper describes the initial stages of a project aiming at improving a real ICS, in particular from the interaction point of view, taking into account the stringent usability requirements from the MDs. Through a validated low definition prototype of the infection module of ICS, the paper proposes innovative active ways of providing suggestions to MDs on what actions to take.

Abstract
Markov models have been widely used to represent and analyze user Web navigation data. In previous work, we have proposed a method to dynamically extend the order of a Markov chain model and a complimentary method for assessing the predictive power of such a variable-length Markov chain. Herein, we review these two methods and propose a novel method for measuring the ability of a variable-length Markov model to summarize user Web navigation sessions up to a given length. Although the summarization ability of a model is important to enable the identification of user navigation patterns, the ability to make predictions is important in order to foresee the next link choice of a user after following a given trail so as, for example, to personalize a Web site. We present an extensive experimental evaluation providing strong evidence that prediction accuracy increases linearly with summarization ability.

Abstract
We present two methods for testing the predictive power of a variable length Markov chain induced from a collection of user web navigation sessions. The collection of sessions is split into a training and a test set. The first method uses a chi(2) statistical test to measure the significance of the distance between the distribution of the probabilities assigned to the test trails by a Markov model build from the full collection of sessions and a model built from the training set. The statistical test measures the ability of the model to generalise its predictions to the unseen sessions from the test set. The second method evaluates the model ability to predict the last page of a navigation session based on the preceding pages viewed by recording the mean absolute error of the rank of the last occurring page among the predictions provided by the model. Experimental results conducted on both real and random data sets are reported and the results show that in most cases a second-order model is able to capture sufficient history to predict the next link choice with high accuracy.

Abstract
This paper describes how Fuzzy Logic helped the representation and handling of equipment fault descriptions and diagnosis in a Case-Based Reasoning (CBR) system, in a way close to common spoken language, frequently vague, used by maintenance teams in daily maintenance tasks. These descriptions may be of two types as sometimes they contain a judgment based on the technician own know-how and sometimes they don't. For case similarity computation the "meaning" of these observations must be understood and all must be reduced to the same format before any further calculus may take place. In this context another important issue is relevance: the weight of each observation for each possible diagnosis is automatically captured by the CBR system according to the experiences it stores along its life-cycle. A prototype has been tested in the health-equipment maintenance field. The system acts as a know-how repository and a divulgation support, part Of an intelligent e-learning plafform for staff training,

Abstract
One of the issues in the shoe industry is that market seeks the fulfilment of product orders for smaller production lots with shorter deadlines. A need therefore arises to respond to these market changes. Such a response includes amongst other things improving the internal organization of the enterprise, by for example rationalizing the production floor layout. A portuguese shoe factory decided to cut costs related to the transportation of all work in progress within the factory plant by implementing a fully automated storage and retrieval system. This Paper presents the development of a Simulation model, using the Arena simulation software tool, to analyse the performance of the automated storage and retrieval system based oil a number of scenarios. The simulation model uses VBA (Visual Basic for Applications) to implement the control logic and graphics to Visualize the sequences of movements and storage / retrieval operations.