China
Africa
Explore chapters and articles related to this topic
Semantic Technologies as Enabler
Published in Sarika Jain, Understanding Semantics-Based Decision Support, 2021
RDF4J (Sesame) is an official fork of the OpenRDF Sesame project. OpenRDF Sesame is a Java-based open-source framework for storage, inference, and querying of RDF and RDF Schema. It can be used as a stand-alone server—i.e., a database for RDF and RDF Schema—and also as a Java library. Sesame is a lightweight yet powerful API with reasoning and transactional support.Storage: Eclipse RDF4J, or Sesame, is an open-source, database-independent storage for RDF data. It can be combined with a variety of DBMSs like MySQL, PostgreSQL, or Oracle 9i or newer. In its architecture it contains a layer named SAIL (Storage and Inference Layer) for managing communication with the database in use. It provides support for all three types of storage: in-memory and native data stores and APIs for integrating with RDF databases.Query: Sesame can accept only queries written in SeRQL (an RDF query language) and converts them into queries suitable to run on the underlying repository.Reasoning: It does not have any support for OWL, so it is not suitable for ontologies.14
Considerations on the Use of Custom Accelerators for Big Data Analytics
Published in Kuan-Ching Li, Hai Jiang, Albert Y. Zomaya, Big Data Management and Processing, 2017
Vito Giovanni Castellana, Antonino Tumeo, Marco Minutoli, Marco Lattuada, Fabrizio Ferrandi
The emerging data models, the limitations of old frameworks, and the issues of commodity systems have led to the introduction of countless specialized software infrastructures. Open source RDF databases include Jena SDB [1] backed by relational databases; Jena TDB [1] backed by native, disk-based storage; Sesame [2] with support for layering on top of relational databases or a native backend; Virtuoso Open Source edition [3]; and 4store [4]. Research-level RDF database approaches include RDF-3X [5], Hexastore [6], YARS2 [7], SHARD [8], BitMat [9], and SPARQL queries on PIG [10]. Commercial RDF databases include Bigdata [11], BigOWLIM [12], and Virtuoso [3]. A number of these approaches leverage Map-Reduce frameworks to achieve scalability on clusters, or implement support for distributed cluster with the objective to increase performance, dataset size, or both [13–15]. The SPARK framework [16], on top of Hadoop or in standalone mode, together with the GraphX library, has also started to be employed to implement and query RDF databases. GEMS, the Graph Engine for Multithreaded Systems [17], is a software stack composed of three layers: a custom multithreaded runtime (GMT—Global Memory and Threading), a library of graph methods and related data structures, and a SPARQL-to-C++ compiler. The custom runtime at its basis enables irregular applications and, more specifically, graph methods, to scale in performance and size on commodity clusters by addressing through software approaches some of the shortcomings of current high-performance architectures with this type of applications. Specifically, it provides a global address space across cluster nodes (so that large datasets to be explored in memory do not need partitioning), lightweight software multithreading to tolerate latencies to access data on remote nodes, and messages aggregation to maximize utilization of the interconnect network of the cluster. The SPARQL-to-C++ compiler converts SPARQL queries into graph methods expressed in C++ that exploit the graph library at the lower level of the stack. GEMS is a first example of how hardware characteristics of systems need careful consideration for emerging big data analytics workloads. The original Cray Urika is an example of solutions that tried to go even further on the path of specialization. It employed the custom-designed Cray XMT 2 architecture (derived from the Tera MTA designs, systems implementing nodes with multithreaded processors, global address space, and synchronization at the level of the single memory word to better support throughput-based applications such as graph exploration) to implement graph databases, adapting existing open source frameworks. Cray has moved away from the use of custom architectures due to maintainability and cost issues, and introduced a Urika-GX [18], based on somewhat more commodity components (x86 nodes with the latest Cray Aries interconnect) and running Hadoop and Spark software frameworks. However, recognizing that these are not optimal for all type of workloads, it also supports the Cray Graph Engine (CGE), which again employs a custom runtime with features that can speed up graph methods.
Design for invention: a framework for identifying emerging design–prior art conflict
Published in Journal of Engineering Design, 2018
Pingfei Jiang, Mark Atherton, Salvatore Sorce, David Harrison, Alessio Malizia
The generated RDF files were then uploaded to an Eclipse RDF4J server (RDF4J n.d.). RDF4J is an open-source framework, formerly known as Sesame, for querying and analysing RDF data. In our case, we deployed it over an instance of an Apache-Tomcat web server. The RDF4J server can then be accessed both from a web interface (browser-based access) and from a URI (Uniform Resource Identifier – for programmatic access), both for querying and managing. The server accepts queries in different languages, and in our study SPARQL Protocol and RDF Query Language was used due to its broad application and popularity. SPARQL enables the designer to describe the working principles of an emerging design in the form of one or multiple queries such that possible matches from the database can be retrieved. For instance, in these queries the designer can specify emerging design geometric features, FGI or multiple FGI. By doing this the designer is able to obtain information of any potential conflicted prior art. Figure 7 illustrates an overview of the patent data coding process, starting from the patents in their common form (e.g. PDF), to their representation in RDF along with supplementary semantic data, in order to allow SPARQL querying for potential conflicts. In the next section several example queries were conducted to demonstrate the emerging design–prior art comparison method.