The realm of big data has brought new venues for knowledge acquisition, but also major challenges including data interoperability and effective management. The great volume of miscellaneous data renders the generation of new knowledge a complex data analysis process. Presently, big data technologies provide multiple solutions and tools towards the semantic analysis of heterogeneous data, including their accessibility and reusability. However, in addition to learning from data, we are faced with the issue of data storage and management in a cost-effective and reliable manner. This is the core topic of this paper. A data lake, inspired by the natural lake, is a centralized data repository that stores all kinds of data in any format and structure. This allows any type of data to be ingested into the data lake without any restriction or normalization. This could lead to a critical problem known as data swamp, which can contain invalid or incoherent data that adds no values for further knowledge acquisition. To deal with the potential avalanche of data, some legislation is required to turn such heterogeneous datasets into manageable data. In this article, we address this problem and propose some solutions concerning innovative methods, derived from a multidisciplinary science perspective to manage data lake. The proposed methods imitate the supply chain management and natural lake principles with an emphasis on the importance of the data life cycle, to implement responsible data governance for the data lake.

International audienceEnvironmental resources (e.g., air quality, water quantity) are needed to understand fundamental questions such as global change. Such resources are often collected from sensors, including humans acting as sensors. Tools have emerged to manage such data in the form of time series and, in particular, the Sensor Observation Service (SOS) which offers a framework based on predefined relational database schema. Environmental observatories can be built using such frameworks, allowing to address specific key scientific questions by collecting and sharing large-scale environmental data. However, the strict schema of SOS database makes it difficult to integrate some data that cannot be directly mapped to the schema. Guidelines and best practices are offered in the literature in order to reuse standards from the Semantic Web but they do not cover all needs. In particular, they do not help to reflect the fact that a single environmental database can lead to several SOS models. Since being aware of these multiple possibilities is crucial for a better use of the observatories, we argue that some extensions of the existing works are required. In this paper, we thus propose an extension of existing vocabularies to achieve this goal. Our contribution is illustrated on the real case of the Lebanese-French O-LiFE environmental observatory

No abstract

Environmental data are currently gaining more and more interest as they are required to understand global changes. In this context, sensor data are collected and stored in dedicated databases. Frameworks have been developed for this purpose and rely on standards, as for instance the Sensor Observation Service (SOS) provided by the Open GeoSpatial Consortium (OGC), where all measurements are bound to a so-called Feature of Interest (FoI). These databases are used to validate and test scientific hypotheses often formulated as correlations and causality between variables, as for instance the study of the correlations between environmental factors and chlorophyll levels in the global ocean. However, the hypotheses of the correlations to be tested are often difficult to formulate as the number of variables that the user can navigate through can be huge. Moreover, it is often the case that the data are stored in such a manner that they prevent scientists from crossing them in order to retrieve relevant correlations. Indeed, the FoI can be a spatial location (e.g., city), but can also be any other object (e.g., animal species). The same data can thus be represented in several manners, depending on the point of view. The FoI varies from one representation to the other one, while the data remain unchanged. In this article, we propose a novel methodology including a crucial step to define multiple mappings from the data sources to these models that can then be crossed, thus offering multiple possibilities that could be hidden from the end-user if using the initial and single data model. These possibilities are provided through a catalog embedding the multiple points of view and allowing the user to navigate through these points of view through innovative OLAP-like operations. It should be noted that the main contribution of this work lies in the use of multiple points of view, as many other works have been proposed for manipulating, aggregating visualizing and navigating through geospatial information. Our proposal has been tested on data from an existing environmental observatory from Lebanon. It allows scientists to realize how biased the representations of their data are and how crucial it is to consider multiple points of view to study the links between the phenomena.