We propose a new efficient algorithm for solving the cluster labeling problem in support vector clustering (SVC). The proposed algorithm analyzes the topology of the function describing the SVC cluster contours and explores interconnection paths between critical points separating distinct cluster contours. This process allows distinguishing disjoint clusters and associating each point to its respective one. The proposed algorithm implements a new fast method for detecting and classifying critical points while analyzing the interconnection patterns between them. Experiments indicate that the proposed algorithm significantly improves the accuracy of the SVC labeling process in the presence of clusters of complex shape, while reducing the processing time required by existing SVC labeling algorithms by orders of magnitude.

To reduce the prohibitive efforts associated to problem-solving ITS, a domain-independent framework can be developed. In this paper, we compare ASTUS to the MTT architecture with examples drawn from a scatter plot tutor. We show that both approaches are similar, but that an ASTUS tutor can exhibit a more sophisticated tutoring behavior thanks to its “pedagogical knowledge model”.

Clustering categorical data faces two challenges, one is lacking of inherent similarity measure, and the other is that the clusters are prone to being embedded in different subspace. In this paper, we propose the first divisive hierarchical clustering algorithm for categorical data. The algorithm, which is based on multiple correspondence analysis (MCA), is systematic, efficient and effective. In our algorithm, MCA plays an important role in analyzing the data globally. The proposed algorithm has five merits. First, our algorithm yields a dendrogram representing nested groupings of patterns and similarity levels at different granularities. Second, it is parameter-free, fully automatic and, most importantly, requires no assumption regarding the number of clusters. Third, it is independent of the order in which the data are processed. Forth, it is scalable to large data sets; and finally, using the novel data representation and Chi-square distance measures makes our algorithm capable of seamlessly discovering the clusters embedded in the subspaces. Experiments on both synthetic and real data demonstrate the superior performance of our algorithm.

This paper presents a task selection model for personalised educational instruction. The proposed model is based on the student expertise level and it takes into account performance and mental load in order to adapt task selection. The article also discusses distinctions between our approach and related works.

This paper describes how a knowledge model allows training software to evaluate spatial cognitive maps and provide tailored assistance.

We present in this paper an intelligent tutoring system using a Bayesian network. This tutor is dedicated to the analysis and diagnosis of student's errors. The elaboration of such a system necessitates nearly always taking into consideration information that is potentially incomplete or uncertain. Indeed, in a learning situation, we can neither know exactly the student's plan nor his goal. In addition, we cannot observe what the student knows or does not know, but we can only make imperfect estimations through his actions. In order to model the student in this situation, we designed and implemented an intelligent system that uses Bayesian networh.

In this paper we present an authoring tool milieu that permits modelling graphically any subject-matter domain knowledge and transposing it automatically into related XML files. Generated contents serve as a tutor reasoning support when interacting with students engaged in learning activities through virtual learning environments.

We present some characteristics of a new computational model to represent knowledge within intelligent tutoring systems. This model is inspired (1) by research in artificial intelligence on modelling, structuring and organising knowledge and (2) by theories of cognitive psychology which explain the human cognitive activity in terms of memory subsystems and their processes. Here, we describe one originality of our novel approach: the parsimonious use of cognitive structures suggested by psychology to encode the knowledge to be taught. Primitive units of semantic and procedural knowledge, chosen with a small level of granularity, are used to build complex knowledge entities which are dynamically combined in order to represent the learner cognitive activity. Traces of this activity are used as episodic knowledge which is specific to each learner. Episodic knowledge analysis allows to retrieve semantic and procedural knowledge and to establish their acquisition level in function of their context of usage.