Wildfire prediction has drawn a lot of researchers’ interest, but still presents a computational difficulty since it necessitates real-time data collected from several distributed data sources. Furthermore, because environmental Web services have, now, access to a wider range of environmental data sources, services might be functionally similar but of varying quality. In this paper, we propose a knowledge-driven framework for service composition that is based on a layered architecture. Based on these layers, the proposed framework aims to select the optimal service instances participating in a service composition schema, through a modular ontology to infer the quality of data sources (QoDS) and an outranking approach. Moreover, it aims to executing the service composition schema at runtime by dynamically readjusting both the service composition schema and the service instances via a machine learning-based service composition approach. The conducted experiments showed that the proposed framework enables (i) a reasonable reasoning time for assessing the data sources’ quality, (ii) a decrease in the ELECTRE III MCDM method’s execution time achieved by combining the skyline and adominance methods, (iii) dynamic generation of the most relevant service composition schema with the appropriate wildfire risk classes, and (iv) a high prediction accuracy using our proposed outranking approach compared to the randomly selected services.
Temporal data (TD) in Semantic Web are affected by different types of imperfections principally conflict. In the literature, most of the proposed approaches deal with perfect TD. However, to our knowledge, there is no approach to dealing with conflicting TD. In this paper, we propose an approach to represent and reason about quantitative conflicting TD (i.e., time intervals and points) and associated qualitative relations (e.g., “before”). This approach is based on evidence theory and it is three folds. (i) For the representation, the mass function of the conflicting temporal data is estimated, through the believability measures estimated based on our previous DBE_ALZ approach. Then, an ontology-based representation for the handled TD associated with the obtained mass function is proposed. (ii) For the reasoning, our approach relies on the Allen’s interval relations. First, we extend this algebra to reason about conflicting temporal relations. The resulting interval relations preserve the properties of the original algebra. Second, we adapt the proposed relations to define new ones relating a time interval and a time point, and two time points. All the proposed relations can be used for temporal reasoning through transitivity tables. (iii) Based on (i) and (ii), we propose a new evidential ontology named “BeliefTimeOnto”. We implement a prototype to ease the interaction with the proposed ontology. We conduct two case studies: the first is about temporal data entered by Alzheimer’s patients in the context of a memory prosthesis and the second is about data entered the context of Collective Memory application. The evaluation proves the usefulness of the proposed approach as all the inferences are well established and the precision results are interesting.
Les données temporelles fournies par les patients atteints Alzheimer sont sujettes à l’incertitude. De nombreuses approches ont été proposées pour traiter des données temporelles certaines, mais non pas celles qui sont incertaines. Cet article propose une approche pour représenter et raisonner sur des intervalles et des points de temps quantitatifs certains et incertains et les relations qualitatives entre eux. Elle inclut trois volets : (1) une extension de l’approche 4D-fluents avec de nouvelles composantes ontologiques pour représenter des données temporelles certaines et incertaines. (2) une extension de l’algèbre des intervalles d’Allen pour raisonner sur des intervalles de temps certains et incertains. Une adaptation de relations cette algèbre pour relier un intervalle de temps et un point de temps, et deux points de temps. (3) Enfin une ontologie qui intègre toutes ces extensions. Un prototype a été implémenté et intégré dans une prothèse de mémoire pour les patients atteints de la maladie d’Alzheimer afin de gérer des entrées de données incertaines.
Dealing with imperfect temporal data entries in the context of Collective and Personal Memory applications is an imperative matter. Data are structured semantically using an ontology called “Collective Memo Onto”. In this paper, we propose an approach that handles temporal data imperfections in OWL 2. We reduce to four types of imperfection defined in our typology of temporal data imperfections which are imprecision, uncertainty, simultaneously uncertainty and imprecision and conflict. The approach consists of representing imperfect quantitative and qualitative time intervals and time points by extending the 4D-fluents approach and defining new components, as well as reasoning about the handled data by extending the Allen’s Interval algebra. Based on both extensions, we propose an OWL 2 ontology named “TimeOntoImperfection”. The proposed qualitative temporal relations are inferred via a set of 924 SWRL rules. We validate our work by implementing a prototype based on the proposed ontology and we apply it in the context of the Collective Memory Temporal Data.