The increase in the number of complex temporal datasets collected today has prompted the development of methods that extend classical machine learning and data mining methods to time-series data.  This work focuses on methods for multivariate time-series classification. Time series classification is a challenging problem mostly because the number of temporal features that describe the data and are potentially useful for classification is enormous. We study and develop a temporal abstraction framework for generating multivariate time series features suitable for classification tasks. We propose the STF-Mine algorithm that automatically mines discriminative temporal abstraction patterns from the time series data and uses them to learn a classification model. Our experimental evaluations, carried out on both synthetic and real world medical data, demonstrate the benefit of our approach in learning accurate classifiers for time-series datasets.

Evaluating changes in documentation of technical products is a key issue in knowledge management. A product may be declined in different versions and one way to evaluate changes is to compare the sets of documents which describe each version. The aim of this paper is to propose a framework for exhibiting changes between sets of documents. This framework is based on the representation of the sets of documents in terms of association rules and on the definition of first order predicates for reasoning with these association rules. The aim of the reasoning stage is to exhibit the differences between the sets of documents. These predicates show what rules are specific to a corpus or how differs the usage of concepts appearing in the associations rules. The framework is  experimented with the comparison of two corpuses of documents which describe documentation about two different versions of a spatial component.

Our past work has investigated the use of the Cognitive Model Software Development Kit (SDK) for creating the cognitive models that underlie model-tracing Cognitive Tutors. Though successful at increasing the number of people who could author such a cognitive model, for certain kinds of situations the Cognitive Model SDK proved cumbersome. The present work discusses a new authoring system, xPST, that allows an example-based tutor to be built on top of existing software.

Boosting is a well-known method for improving the accuracy of many learning algorithms. In this paper, we propose a novel boosting algorithm, VipBoost (voting on boosting classifications from imputed learning sets), which first generates multiple incomplete datasets from the original dataset by randomly removing a small percentage of observed attribute values, then uses an imputer to fill in the missing values.  It then applies AdaBoost (using some base learner) to produce classifiers trained on each of the imputed learning sets, to produce multiple classifiers. The subsequent prediction on a new test case is the most frequent classification from these classifiers. Our empirical results show that VipBoost produces very effective classifiers that significantly improve accuracy for unstable base learners and some stable learners, especially when the initial dataset is incomplete.

We present a semantically-driven approach to uncertainties within and across ontologies. Ontologies are widely used not only by the Semantic Web but also by artificial systems in general. They represent and structure a domain with respect to its semantics. Uncertainties, however, have been rarely taken into account in ontological representation, even though they are inevitable when applying ontologies in `real world' applications. In this paper, we analyze why uncertainties are necessary for ontologies, how and where uncertainties have to be represented in ontologies, and what their semantics are. In particular, we investigate which ontology constructions need to address uncertainty issues and which ontology constructions should not be affected by uncertainties on the basis of their semantics. As a result, the use of uncertainties is restricted to appropriate cases, which reduces complexity and guides ontology development. We give examples and motivation from the field of spatially-aware systems in indoor environments.

This paper introduces a qualitative ranking function that uses signed integers to describe the surprise associated with the occurrence of events. The measure introduced, kappa++, is based on the kappa calculus but differs from it in that its semantics enable an explicit representation of complements. As a result, the kappa++ is more capable of enforcing probability theory-like constraints to carry on reasoning.

This volume contains the papers presented at the 22nd International FLAIRS Conference (FLAIRS-22) held 19-21 May 2009 on Sanibel Island, Florida, USA. The call for papers attracted 158 paper submissions, 40 to the general conference and 118 to the 10 special tracks. Over 80 percent of the papers were reviewed by at least four reviewers, and all papers by at least three. Reviewing was coordinated by the program committees of the general conference and the special tracks. The program committees finally accepted the 85 papers that appear in these proceedings, all as presented papers (21 from the general conference and 64 from the special tracks) and 29 as poster papers (6 from the general conference and 23 from the special tracks).

In our experiments with four well-known systems for solving partially observable planning problems  (Contingent-FF, MBP, PKS, and POND), we were greatly surprised to find that they could only solve problems with a small number of contingencies. Apparently they were repeatedly trying to solve many combinations of contingencies at once, thus unnecessarily using up huge amounts of time and space. This difficulty can be alleviated if the planner can maintain focus on the contingency that it is currently trying to solve. We provide a way to accomplish this by incorporating focusing information directly into the planning domain's operators, without any need to modify the planning algorithm itself. This enables the above planners to solve larger problems and to solve them much more quickly. We also provide a new planner, FOCUS, in which focusing information can be provided as a separate input. This provides even better performance by allowing the planner to utilize more extensive focusing information.

We propose a new approach for Temporal Diagnosis Problems. This approach is an extension of  Bouzid and Ligeza's  method for temporal diagnosis problems. In this latter work, the authors define a Temporal Causal Graph (TCG) where time delays are expressed as temporal instants. We extend the TCG by including two quantitative relations in order to handle temporal intervals. We call ExTCG this new model. Solving a temporal diagnosis problem represented by the ExTCG consists of finding all possible explanations. It is performed using a backtrack search algorithm. In many diagnosis applications, the generation of all possible explanations is not necessary. For this reason, we augment the ExTCG in order to consider the degree of causality between symptoms. We call weighted ExTCG this extended model. Solving it consists of finding the explanation  having the highest probability to occur. Through a real world diagnosis application in medicine, we illustrate the weighted ExTCG and its corresponding solving algorithm.

Scientists typically need to take a large volume of information into account in order to deal with re-occurring tasks such as inspecting proceedings, finding related work, or reviewing papers. Our work aims at filling the gap between text documents and a structured representations of their content in the domain of resilience computing by combining computer linguistics and ontological methods. The results of our research include: a thesaurus of the domain, automatic clustering of the domain documents, a domain ontology, and a tool for constructing ontologies with the aid of domain thesauri.