Thomas Raimbault

Big data: A new challenge for tourism

This paper proposes the foundations for a definition of digital tourist survey field based on the study of social networks. Several social network provide essential information on the perceptions of tourist destination. Nevertheless to operate correctly, several points should be discussed. This article provides multiple ideas for reflections on the challenges and opportunities offered by these new media.

Overviewing the RDF(S) Semantic Web

The RDF(S) semantic web consists of many billions statements about resources (RDF data and RDFS metadata). We propose an original and effective approach to improve performance of retrieving information in RDF(S) statements sets, by quickly discarding irrelevant RDF data sets and moreover using classic search ways (like SPARQL and Jena). To that end, we define and associate at each RDF data set a kind of overview, small and search-oriented, which has to be queried first before querying whole data.

Conceptual Graphs and Datatypes

Datatypes , like numbers or strings, are widely used in Knowledge Representation (e.g. in RDF(S)/OWL or UML languages). The usual model of simple conceptual graphs does not support datatypes. Some extensions of conceptual graphs have been proposed for using datatypes, however these extensions often wander from initial model of conceptual graphs by introducing for instance procedural relations between nodes. This paper proposes a datatype extension for the simple conceptual graph model. Our contribution is threefold. First, we allow the use of datatypes for typing concept nodes. Second, we define two families of conceptual graphs: factual graphs and query graphs , both close to initial model. Factual graph is used to represent factual knowledge, including values of datatypes. Query graph may contain concept nodes that represent conditional queries on values of datatypes; these conditions are expressed by regular operators on datatypes. Third, we adapt projection to operate from a query graph to a factual graph.

A new method to interrogate and check UML class diagrams

We present a new method for graphically interrogating and checking UML class diagrams. We employ the model of conceptual graphs (CGs) as representation, calculation and visualisation model. The key idea of our work is to translate UML class diagrams into the formalism of CGs. First, UML notations are encoded into UML Ontology that is a support of CG. Second, using the UML Ontology, a UML class diagram can be translated into a CG, called CG class diagram. Third, CG class diagrams can be interrogated via the elementary operation of CG, named projection. Fourth, constraints and rules provides a way to model specifications for checking CG class diagrams. We use two approaches to check a class diagram: object-oriented specifications and field specifications.

Consistency Evaluation of RDF Data: How Data and Updates are Relevant

Trust and quality maintenance have always been problematic in the Semantic Web RDF bases. Numerous propositions to address these problems of data integration have been made, either based on ontologies or on additional metadata. However ontologies suffer from a adaptation speed slower than the data evolution speed and metadata requires ad-hoc manipulations of data by addition of extra-data. In this article we propose an original approach, based exclusively on data from the base, to evaluate the consistency of a candidate update to a RDF base, and finally to know if this update is relevant to the base. Our approach is inspired by case-based reasoning and uses similarity evaluation and query relaxation methods to compare a candidate update to the data from the base. If the modifications of a candidate update make the target part of the base more similar to other part (s) of the base, then this candidate update is considered consistent with the base and can be applied.

Targeted Linked-Data Extractor

This paper proposes a method to extract smartly a sub-part of the Linked Data Cloud starting with a list of resources as seeds. This method allow us to extract knowledge starting from a selected list of resources and expanding it to their neighbours.

Cognitive map of tourist behavior based on Tripadvisor

The objective of this paper is to identify, based on data from Tripadvisor, tourist behavior in how users rate a tourist place. Firstly, we propose different correspondence analyses (CA) on data from Tripadvisor to discover pairwise dependences between data properties (e.g. rating vs. place type, age vs. country). Secondly, we merge and map all our CA results as a cognitive map, both to bring out and understand influences between the studied concepts and to make easier human visualization.

Aperçus de recherche : interroger efficacement un ensemble de bases RDF

Dans le cadre du web semantique, nous proposons une approche de recherche d’information basee sur une version compacte d’une base de triplets RDF, appelee apercu de recherche, agissant comme une vue d’ensemble de cette base. Les apercus de recherche permettent d’ecarter rapidement et de facon sure une base du spectre des recherches si celle-ci ne contient pas d’elements satisfaisant a une requete donnee. L’interet d’utiliser un apercu est – d’un point de vue performance – qu’il n’est pas necessaire d’interroger la base associee toute entiere et – d’un point de vue pratique – que l’interrogation d’un apercu passe par des moyens classiques SPARQL. Nous avons evalue notre approche sur des donnees RDF reelles, extraites de DBPedia, selon un jeu de requetes choisies parmi celles couramment envoyees a DBPedia.

Transitions-based system: A relevant use case

A transitions-based system is a promising way, introduced in 2012 by Raimbault et al., for simply and effectively using the complementarity between several Knowledge Representation and Reasoning approaches. From a transitions-based system composed of OWL, Conceptual Graphs and UML approaches, we present in this paper a relevant use case to illustrate the advantage of using a such stimulant system. We firstly use this system to represent knowledge from the Paris-Dakar rally, and secondly to reason on that represented knowledge by querying, by checking and by deducing.

USING UML CLASS DIAGRAM AS A KNOWLEDGE ENGINEERING TOOL

UML class diagram is the de facto standard, including in Knowledge Engineering, for modeling structural knowledge of systems. Attaching importance to visual representation and based on a previous work, where we have given a logical defined extension of UML class diagram to represent queries and constraints into the UML visual environment, we present here how using the model of conceptuals graphs to answer queries and to check constraints in concrete terms.