Résumé: The business intelligence and decision-support systems used in many application domains casually rely on data warehouses, which are decision-oriented data repositories modeled as multidimensional (MD) structures. MD structures help navigate data through hierarchical levels of detail. In many real-world situations, hierarchies in MD models are complex, which causes data aggregation issues, collectively known as the summarizability problem. This problem leads to incorrect analyses and critically aects decision making. To enforce summarizability, existing approaches alter either MD models or data, and must be applied a priori, on a case-by-case basis, by an expert. To alter neither models nor data, a few query-time approaches have been proposed recently, but they only detect summarizability issues without solving them. Thus, we propose in this paper a novel approach that automatically detects and processes summarizability issues at query time, without requiring any particular expertise from the user. Moreover, while most existing approaches are based on the relational model, our approach focus on an XML MD model, since XML data is customarily used to represent business data and its format better copes with complex hierarchies than the relational model. Finally, our experiments show that our method is likely to scale better than a reference approach for addressing the summarizability problem in the MD context.

Ref HAL: hal-00728197_v1
Résumé: In this vision paper, we discuss models and techniques for integrating, processing and querying data, information and knowledge within data warehouses in a user-centric manner. The user-centric emphasis allows us to achieve a number of clear advantages with respect to classical data warehouse architectures, whose most relevant ones are the following: (i) a unified and meaningful representation of multidimensional data and knowledge patterns throughout the data warehouse layers (i.e., loading, storage, metadata, etc); (ii) advanced query mechanisms and guidance that are capable of extracting targeted information and knowledge by means of innovative information retrieval and data mining techniques. Following this main framework, we first outline the importance of knowledge representation and management in data warehouses, where knowledge is expressed by existing ontology or patterns discovered from data. Then, we propose a user-centric architecture for OLAP query processing, which is the typical applicative interface to data warehouse systems. Finally, we propose insights towards cooperative query answering that make use of knowledge management principles and exploit the peculiarities of data warehouses (e.g., multidimensionality, multi-resolution, and so forth).

Ref HAL: hal-00727331_v1
DOI: 10.1145/2347673.2347683
Résumé: Business intelligence (BI) is a broad field related to integrating, storing and analyzing data to help decision-makers in many domains (from actual business fields to administration, health and environment) make better decisions. Front-end analytics methods include reporting, on-line analytical processing (OLAP), and data mining. With the increasing success of cloud computing, cloud business intelligence “as a service” offerings have sparkled widely, both from cloud start-ups and major BI industry vendors. Beyond porting BI features into the cloud, which already implies numerous issues (e.g., BigData/NoSQL database modeling and storage, data localization, security and privacy, performance, cost and usage models…), this trend also poses new, broader challenges for making data analytics available to small and middle-size enterprises (SMEs), non-governmental organizations, web communities (e.g., supported by social networks), and even the average citizen; this vision presumably requiring a mixture of both private and open data. Thus, Cloud Intelligence is not only a current technological and research challenge, but also an important societal stake, since people increasingly demand open data (e.g., the Spanish indignados), which they need to access easily from the Web, to possibly mix them with private data, to analyze them with intelligible on-line tools with advanced collaborative features that enable users to share and re-use business intelligence concepts and analyses in large scale fashion, and to share the results world-wide. The aim of this workshop is to become an interdisciplinary, regular exchange forum for researchers, industry and practitioners, as well as all potential users of Cloud Intelligence. The submission of research, industrial, position, visionary and student papers are encouraged to fuel up the discussion.

Commentaires: ISBN: 978-1-4503-1596-8 http://dl.acm.org/citation.cfm?id=2347683

Résumé: We propose an original approach for mapping the choreographic data into a new representation language adapted to data mining techniques. Our approach relies mainly on the notion of “dance tags” that we took from the NLP community by analogy with Part-of-Speech tagging. The process starts from scores described in Labanotation and produces in a fully automatic manner a high-level, comprehensive representation of the choreographic sequence. Our experiments show that we succeed in retrieving manually translated scores with an accuracy of 85% to 94%. Using this new representation of the choreographic data, one can then perform several useful tasks in an efficient manner. Among these are: music recommendation, automated detection of dance style or genre, and ultimately any task that requires a deeper understanding of the meaning of choreographic information than traditional processing can provide. In this paper, we demonstrate the usefulness of our approach with a simple example for discriminating between classical ballet, modern ballet, and folkloric dances.

Résumé: Abstract—In this paper, we propose a new time-aware dissimilarity measure that takes into account the temporal dimension. Observations that are close in the description space, but distant in time are considered as dissimilar. We also propose a method to enforce the segmentation contiguity, by introducing, in the objective function, a penalty term inspired from the Normal Distribution Function. We combine the two propositions into a novel time-driven constrained clustering algorithm, called TDCK-Means, which creates a partition of coherent clusters, both in the multidimensional space and in the temporal space. This algorithm uses soft semi-supervised constraints, to encourage adjacent observations belonging to the same entity to be assigned to the same cluster. We apply our algorithm to a Political Studies dataset in order to detect typical evolution phases. We adapt the Shannon entropy in order to measure the entity contiguity, and we show that our proposition consistently improves temporal cohesion of clusters, without any significant loss in the multidimensional variance.

---------

Résumé: Online discussions became increasingly widespread with the Web 2.0: no matter the distance, whether you know the person or not, you can discuss and exchange ideas with people all over the world through forums, blogs, and newsgroups. The news websites have extensively used forums in order to encourage the reader being a real participant in the information media. This paper aims at automatically extracting the celebrities from such discussions. We propose certain meta-criteria and we provide an evaluation on a dataset of 35,175 posts written by 14,443 users. The results show that one of the proposed meta-criteria succeeds in extracting celebrities and allows for further improvements.

Résumé: Business Intelligence plays an important role in decision making. Based on data warehouses and Online Analytical Processing, a business intelligence tool can be used to analyze complex data. Still, summarizability issues in data warehouses cause ineffective analyses that may become critical problems to businesses. To settle this issue, many researchers have studied and proposed various solutions, both in relational and XML data warehouses. However, they find difficulty in evaluating the performance of their proposals since the available benchmarks lack complex hierarchies. In order to contribute to summarizability analysis, this paper proposes an extension to the XML warehouse benchmark (XWeB) with complex hierarchies. The benchmark enables us to generate XML data warehouses with scalable complex hierarchies as well as summarizability processing. We experimentally demonstrated that complex hierarchies can definitely be included into a benchmark dataset, and that our benchmark is able to compare two alternative approaches dealing with summarizability issues.