Search

This vision paper lays the preliminary foundations for Data Narrative Management Systems (DNMS), systems that enable the storage, sharing, and manipulation of data narratives. We motivate the need for such formal foundations and introduce a simple logical framework inspired by the relational model. The core of this framework is a Data Narrative Manipulation Language inspired by the extended relational algebra. We illustrate its use via examples and discuss the main challenges for the implementation of this vision.

In this paper, we discuss methods to assess the interestingness of a query in an environment of data cubes. We assume a hierarchical multidimensional database, storing data cubes and level hierarchies. We start with a comprehensive review of related work in the fields of studies of human behavior and computer science. We define the interestingness of a query as a vector of scores along difference dimensions, like novelty, relevance, surprise and peculiarity and complement this definition with a taxonomy of the information that can be used to assess each of these dimensions of interestingness. We provide both syntactic (result-independent) checks and extensional (result-dependent) measures and algorithms for assessing the different dimensions of interestingness in a quantitative fashion. We also report our findings on a user study that we conducted, analyzing the significance of each dimension, its evolution over time and the behavior of the study's participants., 86 pages, 21 figures

International audience

International audience; This paper introduces the Traveling Analyst Problem (TAP), an original strongly NP-hard problem where an automated algorithm assists an analyst to explore a dataset, by suggesting the most interesting and coherent set of queries that are estimated to be completed under a time constraint. We motivate the problem, study its complexity, propose a simple heuristic under simplifying assumptions for approximating it, and run preliminary tests to observe the behavior of this heuristic.

International audience; This vision paper introduces several ideas around the optimization of Interactive Data Analysis (IDA) tasks. With an eye on traditional query optimization (QO) in Relational DataBase Management Systems (RDBMS), we suggest that IDA tasks should be specified through high-level statements and optimized globally , particularly by maximizing the number and significance of insights that can be automatically collected for the task. We envision an architecture for IDA and propose in the context of IDA what corresponds to statistics, cost model and execution plan pruning strategy in relational systems. We give elements pertaining to the feasibility of the vision and draw perspectives.

On-demand integration of multiple data sources is a critical requirement in many Big Data settings. This has been coined as the data variety challenge, which refers to the complexity of dealing with an heterogeneous set of data sources to enable their integrated analysis. In Big Data settings, data sources are commonly represented by external REST APIs, which provide data in their original format and continously apply changes in their structure (i.e., schema). Thus, data analysts face the challenge to integrate such multiple sources, and then continuosly adapt their analytical processes to changes in the schema. To address this challenges, in this paper, we present the Metadata Management System, shortly MDM, a tool that supports data stewards and analysts to manage the integration and analysis of multiple heterogeneous sources under schema evolution. MDM adopts a vocabulary-based integration-oriented ontology to conceptualize the domain of interest and relies on local-as-view mappings to link it with the sources. MDM provides user-friendly mechanisms to manage the ontology and mappings. Finally, a query rewriting algorithm ensures that queries posed to the ontology are correctly resolved to the sources in the presence of multiple schema versions, a transparent process to data analysts. On-site, we will showcase using real-world examples how MDM facilitates the management of multiple evolving data sources and enables its integrated analysis.

This document describes the final implementation and the evaluation of the CHOReOS middleware. Evaluation is achieved both via the use of the middleware on CHOReOS use-cases and via synthetic experiments and simulation. The conclusion was that the implementation of the CHOReOS middleware has achieved a good level of maturity for an open source project and it is ready to be used in real-world, complex choreographies.

This is Part b of Deliverable D1.4, which specifies the final CHOReOS architectural style, that is, the types of components, connectors, and configurations that are composed within the Future Internet of services, as enabled by the CHOReOS technologies developed in WP2 to WP4 and integrated in the WP5 IDRE. The definition of the CHOReOS architectural style is especially guided by the objective of meeting the challenges posed by the Future Internet, i.e.: (i) the ultra large base of services and of consumers, (ii) the high heterogeneity of the services that get composed, from the ones offered by tiny things to the ones hosted on powerful cloud computing infrastructures, (iii) the increasing predominance of mobile consumers and services, which take over the original fixed Inter- net, and (iv) the required awareness of, and related adaptation to, the continuous environmental changes. Another critical challenge posed by the Future Internet is that of security, trust and privacy. However, the study of technologies dedicated to enforcing security, privacy and trust is beyond the scope of the CHOReOS project; instead, state of the art technologies and possibly latest results from projects focused on security solutions are built upon for the development of CHOReOS use cases -if and when needed-. The CHOReOS architectural style that is presented in this deliverable refines the definition of the early style introduced in Deliverable D1.3. Key features of the CHOReOS architectural elements are as follows: (1) The CHOReOS service-based components are technology agnostic and allow for the abstraction of the large diversity of Future Internet services, and particularly traditional Business services as well as Thing-based services; a key contribution of the component formalization lies in the inference of service abstractions that allows grouping services that are functionally similar in a systematic way, and thereby contributes to facing the ULS of the Future Internet together with dealing with system adaptation through service substitution. (2) The CHOReOS middleware-layer connectors span the variety of interaction paradigms, both discrete and continuous, which are used in today's increasingly complex distributed systems, as opposed to enforcing a single interaction paradigm that is commonly undertaken in traditional SOA; a central contribution of the connector formalization is the introduction of a multi-paradigm connector type, which not solely allows having highly heterogeneous services composed in the Future Internet but also having those heterogeneous services interoperating even if based on distinct interaction paradigms. (3) The CHOReOS coordination protocols introduce the third and last type of architectural elements char- acterizing the CHOReOS style. They specifically define the structure and behavior of service-oriented systems within the Future Internet as the fully distributed composition of services, i.e., choreographies; the key contribution of the work lies in a systematic model-based solution to choreography realizability, which synthesizes dedicated coordination delegates that govern the coordination of services.

International audience; The state of the art service search engines allow the users to pick the services they need, based on the quality properties, off ered by these services. To this end, the users should interact with the search engines based on the quality models that are imposed by the engines. This is a signi cant restriction towards making the service-oriented paradigm attractive to the general public. In this paper, we propose an approach that allows a user to specify his perception of quality in terms of a simple, user-de fined quality model. The proposed approach automatically maps the user-defi ned quality model to the search engine's quality model. This mapping forms the basis for ordering, grouping and, in general manipulating, the results of the user's service discovery requests.

The Future Internet envisions a ubiquitous world where available services can be easily discovered and coordinated so as to fit users needs. Service choreographies will play a central role in this vision as an effective means to allow heterogeneous services to suitably collaborate. This deliverable defines the CHOReOS Dynamic Development Process Model by refining and completing the artefacts/ activities and the relationships among them already sketched into the DoW (under the WP2 description), and conceptualized by the the CHOReOS conceptual model [CHO11b]. The process model is an abstract and simplified description of what will be the actual CHOReOS software development process to be defined at M24. The process model describes the "strategy" to be used during the choreography life cycle from design, to development, to maintenance (and hence from static, to runtime, to evolution). The model is made up of activities, common to (almost) every process, but structured in a particular way (i.e., the "CHOReOS way"), hence distinguishing the CHOReOS development process from others.

This deliverable specifies the main concepts of the CHOReOS middleware architecture. Starting from the Future Internet (FI) challenges for scalability, heterogeneity, mobility, awareness, and adaptation that have been investigated in prior work done in WP1, we introduce the aforementioned concepts to deal with the requirements derived from the FI challenges. In particular, we propose an extensible and scalable service discovery approach for the organization and discovery of services that relies on multiple service discovery protocols. Moreover, we introduce an extensible and scalable approach, based on the service bus paradigm, for service access that features the integration and adaptation of multiple interaction protocols. Furthermore, we propose solutions that enable the execution of FI service compositions that range from compositions of choreographed services, developed according to the CHOReOS development process, to massive compositions of things. Finally, we detail the Cloud & Grid middleware facilities that support the overall middleware and the choreographies that are built on it, via a unified API that provides access to multiple cloud infrastructures (e.g., Amazon EC2, HP Open Cirrus, private clouds).

The D1.1 deliverable clarifies baseline, progress, and state of the art that CHOReOS will address. For each of the first four CHOReOS work packages, WP1 to WP4, this deliverable gives a precise definition of the state of the art, an indication of the envisaged progress beyond the state of the art by CHOReOS and the baseline for its research.