Manual knowledge acquisition is usually a costly and time-consuming process. Automatic knowledge acquisition methods can then significantly support the knowledge engineer. In this paper, we propose an approach for rapid knowledge capture. The methodology is based on textual subgroup mining in order to discover dependencies for rule prototyping.

In this paper knowledge representation and inference issues for rule-based systems are discussed. The paper deals with improving the logical calculus of Set Attributive Logic founding an expressive rule language XTT2. Representation extensions are introduced, and practical inference rules provided. The original includes an extended state specification, as well as interpreter design. xamples of rule analysis are given. Visual design tool HQed assuring rule quality is also presented.

In this paper certain knowledge and software engineering methods integration issues are discussed. The principal idea is to consider an effective design and implementation framework for rule design with UML, and implementation with Java. The solution proposed in the paper consists of using a custom knowledge engineering design method for rules in the design stage. The rule base is then transformed to UML behavioral diagrams, which can be considered a visual encoding. The rule implementation involves the serialization to Java language using classes representing the decision tables grouping rules sharing the same attributes.

Recently, the development of distributed knowledge systems has become more attractive due to the existence of new social semantic applications such as semantic knowledge wikis. User-friendly tools like wikis allow for a simple acquisition of formal knowledge, but also pose new challenges in knowledge engineering. In this paper, we reconsider classic criteria for verification in the light of a distributed knowledge base and we discuss novel anomalies that possibly occur during the collaborative development of a distributed knowledge base.

Empirical testing is a very popular evaluation method for the development of intelligent systems. Here, previously solved problems with correct solutions are given as cases to the system. Validity is tested by comparing the expected results with the derived solutions. Besides classic forms of boolean testing of occurring solutions more refined methods are required for a thorough evaluation of real world knowledge systems. We present extended precision and recall functions for interactive knowledge systems that are generalizations of the existing measures. Additionally, we propose a visualization method for inspecting the validation result for interactive systems. A case study with a second-opinion system from the medical domain demonstrates the usefulness of the approach.

Quality management, analysis and evaluation of intelligent systems are important tasks. This paper proposes a data mining approach based on the technique of subgroup mining utilizing a data warehouse that contains data from the respective intelligent system to be evaluated and from other external sources. The context of our work is given by an intelligent documentation and consultation system in the medical domain of sonography. For demonstrating the applicability and benefit of the presented approach, we provide several realworld examples of a case-study applying the approach in the medical domain of sonography.

Instance classification using machine learning techniques has numerous applications, from automation to medical diagnosis. In many problem domains, such as spam filtering, classification must be performed quickly across large datasets. In this paper we begin with machine learning techniques based on the naive Bayes classification and attempt to improve classification performance by taking into account attribute confidence intervals.  Our prediction functions operate over nominal datasets and retain the asymptotic complexity of one-pass learning and prediction functions. We present preliminary results indicating a modest, albeit inconsistent improvement over the naive Bayes classifier alone.

In this search, we focused on supervised learning task using association rules algorithms (association based classification). These algorithms, developed in unsupervised learning, extract all the rules whose the support and confidence exceed a prefixed threshold support. After extracting the frequent itemsets, (i.e their support exceeds the threshold support), algorithms subdivide these itemsets to build the rules, and keep only the rules whose confidence exceeds the threshold confidence. The extraction of class association rules, using these algorithms, have several problems, because of the rules' a posteriori filtering. In the first stage, one extracts useless frequent itemsets, those which do not contain class, whereas the second stage can be simplified, since an itemset containing the class gives place only to only one class rule. In order to be able to work with a low threshold support, we propose FCP-Growth an adaptation of FP-Growth which eliminates the frequent itemsets not containing a class. Moreover, to make the minority class be in advantage during the construction of the class itemsets, we adapt the threshold support, in order to use the same threshold support inside each class.

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.

In this paper, we investigate the hypothesis that plan recognition can significantly improve the performance of a case-based reinforcement learner in an adversarial action selection task. Our environment is a simplification of an American football game. The performance task is to control the behavior of a quarterback in a pass play, where the goal is to maximize yardage gained. Plan recognition focuses on predicting the play of the defensive team. We modeled plan recognition as an unsupervised learning task, and conducted a lesion study. We found that plan recognition was accurate, and that it significantly improved performance. More generally, our studies show that plan recognition reduced the dimensionality of the state space, which allowed learning to be conducted more effectively. We describe the algorithms, explain the reasons for performance improvement, and also describe a further empirical comparison that highlights the utility of plan recognition for this task.