In this paper, we propose an approach to handling uncertain time intervals and related qualitative relations, based on possibility theory. Four contributions are included in this approach. (1) Representing uncertain time intervals and related qualitative relations by extending the 4D-fluents approach with new ontological components. (2) Reasoning about uncertain time intervals by extending the Allen's interval algebra. The resulting interval relations preserve the good properties of the original algebra. (3) Proposing an OWL 2 possibilistic temporal ontology based on 4D-fluents approach extension and Allen's interval algebra extension. The proposed qualitative temporal relations are inferred via a set of SWRL rules. We validate our work by implementing a prototype based on this ontology. (4) Applying our work to PersonLink ontology and comparing the obtained results with our previous works.

Temporal data given by Alzheimer's patients are mostly uncertain. Many approaches have been proposed to handle certain temporal data and lack uncertain ones. This paper proposes an approach to represent and reason about quantitative time intervals and points and qualitative relations between them. It is suitable to handle certain and uncertain temporal data. It includes three parts. (1) The authors extend the 4D-fluents approach with certain components to represent certain and uncertain temporal data. (2) They extend the Allen's interval algebra to reason about certain and uncertain time intervals. They adapt these relations to relate a time interval and a time point, and two time points. All relations can be used for temporal reasoning by means of transitivity tables. (3) They propose a certain ontology based on the extensions. A prototype is implemented and integrated into an ontology-based memory prosthesis for Alzheimer's patients to handle uncertain data inputs. The evaluation proves the usefulness of the approach as all the inferences are well established and the precision results are promising.

In this paper, we propose an ontology-based approach for representing and reasoning about certain and uncertain temporal data. It handles temporal data in terms of quantitative time intervals and points and the qualitative relations between them (e.g., “before”). It includes three parts. (1) We extend the 4D-fluents approach with certain ontological components to represent the handled temporal data in OWL 2. (2) We extend the Allen’s interval algebra to reason about certain and uncertain time intervals. We adapt these relations to allow relating a time interval and a time point, and two time points. All relations can be used for temporal reasoning by means of transitivity tables. (3) The extended Allen’s algebra instantiates the 4D-fluents-based representation. Inferences are based on SWRL rules. Based on this ontology, a prototype is implemented and integrated into an ontology-based memory prosthesis for Alzheimer’s patients to handle certain and uncertain temporal data inputs.

In this paper, we propose an ontology-based approach for representing and reasoning about precise and imprecise time intervals. This approach is three folds: (i) extending the 4D-fluents approach with new crisp and fuzzy components, to represent precise and imprecise time intervals, and temporal relations between them, (ii) extending the Allen's interval algebra to enable reasoning about the handled temporal data, and (iii) creating a Fuzzy-OWL 2 ontology "TimeOnto" that, based on the extended Allen's interval algebra, instantiate our 4D-fluents-based representation. The extension of the Allen's algebra handles precise and imprecise time intervals to express precise (e.g., "Before") and imprecise (e.g., "Just before") temporal relations. Compared to related work, our imprecise relations are personalized, in the sense that they are not limited to a defined set of interval relations and their meanings are determined by the domain expert. For instance, the Allen's relation "Before" may be generalized in 5 imprecise relations, where "Before(1)" means "just before" and gradually the time gap between the two intervals increases until "Before(5)" which means "too long before". To enable this representation, we propose an extension of the Vilain and Kautz's point algebra and a redefinition of the Allen's relations by means of this extended algebra. Unlike most related work, the resulting relations preserve many of the desirable properties of the Allen's algebra. These relations can be used for temporal reasoning by means of a transitivity table. We describe a prototype based on TimeOnto that infers new relations using a set of SWRL and fuzzy IF-THEN rules. This prototype was integrated in an ontology-based memory prosthesis.

International audience

The growth of real-time data generation and stored data leads us to be constantly in thinking about the three V's big data challenges: volume, velocity and variety. Existing RDF Stream Processing (RSP) systems have solved the variety lock by defining a common model for producing, transmitting and continuously querying data in RDF model. On the volume and velocity side, the performances of RSP systems need to be improved particularly in terms of joins process between stored and streaming RDF graphs. Stored RDF data are very important in streaming context (related ontologies, summarized RDF data, non-evolutive RDF data or evolve very slowly over time, etc.) but existing RSP systems such as C-SPARQL, CQELS, SPARQLstream, EP-SPARQL, Sparkwave, etc. use non-optimized and non-scalable approaches for performing join operations between stored and dynamic RDF data. Indeed, these systems need to read the entire local or remote stored RDF data sets while RDF data streams continuously arrived and need to be processed in near real-time. This latency may negatively affect performances in terms of continuous processing and often causes multiple bottlenecks within the network in a distributed environment. That also makes impractical to refresh data or update the stored contents. This paper proposes an approach for distributed real-time joins between stored and streaming RDF graphs using Bloom filters. The join procedure consists of adding fast processing by greatly reducing intermediate results, in-memory indices storage and precomputing query partitions according to the picked SPARQL query variable(s) between the two natures of RDF data. Experimental and evaluations results confirm the performances gained with our approach which significantly speeds up the query processing compared to the actual RSP's techniques.

International audience

In collaborative filtering recommender systems, the users rate items, and this process helps in understanding their preferences. The systems can suffer from the cold-start problem, which refers to the absence or insufficiency of ratings for new users. This can be solved by using active learning strategies, which can be non-personalized or personalized, and which were evaluated and tested previously using different datasets and metrics. In this paper, we present a clearer study by implementing the main non-personalized single-heuristic strategies (random, popularity, co—coverage, variance, entropy, entropy0) on the same dataset, and by evaluating them using the same metrics, in order to have a better comparison. We use the public MovieLens dataset in the experimentations and the results show that the random strategy performs the worst, whereas the entropy0 leads to the best results. All strategies except the random strategy lead to very close results at a certain point, where ratings for almost the same items will have been elicited.

Many studies, in the literature, have affirmed a low level of user satisfaction concerning the understandability and readability of large-scale Linked Data visualizations offered by current available tools. This issue is especially problematic for inexperienced users. To address these requirements, we have extended our previous work Memo Graph, an ontology visualization tool, to provide a user-centered interactive solution for extracting and visualizing Linked Data. It takes aim to provide comprehensible and legible visualization. To manage scalability, it is built on an incremental approach to extract descriptive summarization from a given Linked Data endpoint where it becomes possible to generate a “summary graph” from the most important data (middle-out navigation approach). It offers user interfaces that reduce task complexity for users, especially the inexperienced ones. We tested Memo Graph on a number of Linked Data datasets with encouraging results. We discuss the promising results derived from an empirical evaluation, which affirmed that Memo Graph is useful in visualizing Linked Data and usable.

This paper presents a user-friendly tool, called MEMO GRAPH, for visualizing and navigating ontologies. Compared to related work, MEMO GRAPH is designed to be used by everyone, including ontology experts and users not familiar with ontologies. It provides an accessible and understandable user interface that follows the “design-for-all” philosophy. Precisely, it offers an Alzheimer's patients-friendly interface. The MEMO GRAPH ontology visualization tool is integrated in the CAPTAIN MEMO memory prosthesis and it is applied for visualizing a small-scale ontology (PersonLink) and a large-scale ontology (DBpedia). We discuss the encouraging results derived from the preliminary empirical evaluation, which confirms that MEMO GRAPH is an intuitive and usable ontology visualization tool.

This paper focuses on the items/users representation in the domain of recommender systems. These systems compute similarities between items (and/or users) to recommend new items to users based on their previous preferences. It is often useful to consider the characteristics (a.k.a features or attributes) of the items and/or users. This represents items/users by vectors that can be very large, sparse and space-consuming. In this paper, we propose a new accurate method for representing items/users with low size data structures that relies on two concepts: (1) item/user representation is based on bloom filter vectors, and (2) the usage of these filters to compute bitwise AND similarities and bitwise XNOR similarities. This work is motivated by three ideas: (1) detailed vector representations are large and sparse, (2) comparing more features of items/users may achieve better accuracy for items similarities, and (3) similarities are not only in common existing aspects, but also in common missing aspects. We have experimented this approach on the publicly available MovieLens dataset. The results show a good performance in comparison with existing approaches such as standard vector representation and Singular Value Decomposition (SVD).

Many existing open linked datasets include descriptions of real world persons, with the relationships between them. For some traditional and/or emerging relationships, existing ontologies do not provide the adequate links. This paper represents PersonLink, an ontology that defines rigorously and precisely family relationships, and takes into account the differences that may exist between cultures, including new relationships emerging in our societies nowadays. Moreover, the transition from one culture/language to another one cannot be solved with a simple translation of terms, especially when concepts do not intersect in different languages; thus our solution refers to a multicultural meta-ontology of concepts and associated mechanisms. A validation has been performed on two linked datasets DBpedia and Freebase.

In this paper, we propose a new classification method that addresses classification in multiple categories of textual documents. We call it Matrix Regression (MR) due to its resemblance to regression in a high dimensional space. Experiences on a medical corpus of hospital records to be classified by ICD (International Classification of Diseases) code demonstrate the validity of the MR approach. We compared MR with three frequently used algorithms in text categorization that are k-Nearest Neighbors, Centroide and Support Vector Machine. The experimental results show that our method outperforms them in both precision and time of classification.