Sylvie Salotti

Study and Formalization of a Case-Based Reasoning System Using a Description Logic

In this paper the C-CLASSIC Description Logic is used to design the retrieval and selection tasks of a Case-Based Reasoning system with homogeneous, explicit and formal criteria. The case base is organized by means of a taxonomy of index concepts. Case retrieval is performed using the automatic concept classification of the description logic. Case selection is performed using two criteria: similarity and dissimilarity. Similarity between two cases is characterized by the most specific concept which subsumes the two cases (Least Common Subsumer or LCS), and dissimilarity by a concept representing properties which belong to one case but not to the other. A partial order induced by the subsumption relationship on these concepts is used to select the most similar cases.

CASEP2: Hybrid Case-Based Reasoning System for Sequence Processing

We present in this paper a hybrid neuro-symbolic system called “CASEP2”, which combines the case-based reasoning with an adequate artificial neural network “M-SOM-ART” for sequence classification or prediction task. In CASEP2, we present a new case modelling by dynamic covariance matrices. This model takes into account the temporal dynamics contained in the sequences and allows to avoid problems related to the comparison of different length sequences. In the CBR cycle, one neural network is used during the retrieval phase for indexing the case base and another is used during the reuse phase in order to provide the target case solution.

Case base maintenance for improving prediction quality

We propose a new case base maintenance strategy which allows the improvement of the prediction quality of CBR system. This strategy combines the learning of new selected cases, the reduction of the use of noise cases, and the removal of not useful cases. We describe in this paper the case based reasoning system, in which we have associated new measurements for the source cases. These measurements vary with the use of the cases to get target cases solutions.

A Conceptual Approach for Relational IR: Application to Legal Collections

Everyone who works in the legal field is faced with the complexity of documentary sources of law, that are highly interrelated and interdependent of each others. It is essential for legal practitioners to rely on systems that retrieve all the sources related to the legal cases they are working on and not only the most relevant ones. The challenge for legal IR is to achieve exhaustivity and handle this complexity by retrieving documents on the basis of the semantic content and the intertextual relationships. This work proposes an IR approach for legal sources that goes beyond existing systems. It is based on Formal and Relational Concept Analysis to structure, query and browse collections of legal documents.

Shallow Parsing of INEX Queries

This article presents the contribution of the LIPN : Laboratoire d’Informatique de Paris Nord (France) to the NLQ2NEXI (Natural Language Queries to NEXI) task (part of the Natural Language Processing (NLP) track) of the Initiative for Evaluation of XML Retrieval (INEX 2006). It discusses the use of shallow parsing methods to analyse natural language queries.