Motivated by the recovery and prediction of electricity consumption time series, we extend Nonnegative Matrix Factorization to take into account external features as side information. We consider general linear measurement settings, and propose a framework which models non-linear relationships between external features and the response variable. We extend previous theoretical results to obtain a sufficient condition on the identifiability of NMF with side information. Based on the classical Hierarchical Alternating Least Squares (HALS) algorithm, we propose a new algorithm (HALSX, or Hierarchical Alternating Least Squares with eXogeneous variables) which estimates NMF in this setting. The algorithm is validated on both simulated and real electricity consumption datasets as well as a recommendation system dataset, to show its performance in matrix recovery and prediction for new rows and columns.

Motivated by electricity consumption reconstitution, we propose a new matrix recovery method using nonnegative matrix factorization (NMF). The task tackled here is to reconstitute electricity consumption time series at a fine temporal scale from measures that are temporal aggregates of individual consumption. Contrary to existing NMF algorithms, the proposed method uses temporal aggregates as input data, instead of matrix entries. Furthermore, the proposed method is extended to take into account individual autocorrelation to provide better estimation, using a recent convex relaxation of quadratically constrained quadratic programs. Extensive experiments on synthetic and real-world electricity consumption datasets illustrate the effectiveness of the proposed method.

New personal data fields are currently emerging due to the proliferation of on-body/at-home sensors connected to personal devices. However, strong privacy concerns prevent individuals to benefit from large-scale analytics that could be performed on this fine-grain highly sensitive wealth of data. We propose a demonstration of Chiaroscuro, a complete solution for clustering massively-distributed sensitive personal data while guaranteeing their privacy. The demonstration scenario highlights the affordability of the privacy vs. quality and privacy vs. performance tradeoffs by dissecting the inner working of Chiaroscuro - launched over energy consumption times-series -, by exposing the results obtained by the individuals participating in the clustering process, and by illustrating possible uses.

The advent of on-body/at-home sensors connected to personal devices leads to the generation of fine grain highly sensitive personal data at an unprecendent rate. However, despite the promises of large scale analytics there are obvious privacy concerns that prevent individuals to share their personnal data. In this paper, we propose Chiaroscuro, a complete solution for clustering personal data with strong privacy guarantees. The execution sequence produced by Chiaroscuro is massively distributed on personal devices, coping with arbitrary connections and disconnections. Chiaroscuro builds on our novel data structure, called Diptych, which allows the participating devices to collaborate privately by combining encryption with differential privacy. Our solution yields a high clustering quality while minimizing the impact of the differentially private perturbation. Chiaroscuro is both correct and secure. Finally, we provide an experimental validation of our approach on both real and synthetic sets of time-series.

More and more data mining algorithms are applied to a large number of long time series issued by many distributed sensors. The consequence of the huge volume of data is that data warehouses often contain asynchronous time series, i.e. the values have been sampled and are not anymore observed at the same instants. This is a problem when applying data mining algorithms to such asynchronous time series. The standard way to solve this problem is to interpolate intermediate points. We present here two new interpolation approaches which take into account the knowledge of the integral of the time series between two points. The first approach is naive and uses the history of slope values. The second approach is stochastic and provides a confidence interval of interpolated values. The two methods have been assessed experimentally on a real dataset of electric power consumption time series issued from smart meters.

The growing usage of embedded devices and sensors in our daily lives has been profoundly reshaping the way we interact with our environment and our peers. As more and more sensors will pervade our future cities, increasingly efficient infrastructures to collect, process, and store massive amounts of data streams from a wide variety of sources will be required. Despite the different application-specific features and hardware platforms, sensor network applications share a common goal: periodically sample and store data collected from different sensors in a common persistent memory. In this article we present a clustering approach for rapidly and efficiently computing the best sampling rate which minimizes the SSE (Sum of Square Errors) for each particular sensor in a network. In order to evaluate the efficiency of the proposed approach, we carried out experiments on real electric power consumption data streams produced by a 1-thousand sensor network provided by the French energy group-EDF (Electricite de France).

With the forthcoming deployment of smart metering infrastructures, it will be possible to extract data describing usage and functioning of home devices locally (in residences). Such data can then be streamed continuously and in real-time to a DSMS (data stream management system) for further processing and knowledge extraction. In order for the data to be streamed, their structure and semantics have to be properly defined through a data model. To our knowledge, already existing standards do not include modeling of home devices and their behavior. In this paper, we describe a data stream model that covers these aspects. We also provide basic experimental results based on a data stream generator that we developed and used to test a DSMS.

Automatic clustering methods are part of data mining methods. They aim at building clusters of items so that similar items fall into the same cluster while unsimilar items fall into separate clusters. A particular class of clustering methods are hierarchical ones where recursive clusters are formed to grow a binary tree representing an approximation of similarities between items. We propose a new interactive interface to help the user to interpret the result of such a clustering process, according to the item characteristics. The prototype has been applied successfully to a special case of items providing nice graphical representations electric load curves but can also be used with other types of curves or with more standard items.

Automatic clustering methods are part of data mining methods. They aim at building clusters of items so that similar items fall into the same cluster while dissimilar ones fall into separate clusters. A particular class of clustering methods are hierarchical ones where recursive clusters are formed to grow a tree representing an approximation of similarities between items. We propose a new interactive interface to help the user to interpret the result of such a clustering process, according to the item characteristics. The prototype has been applied successfully to a special case of items providing nice graphical representations (electric load curves) but can also be used with other types of curves or with more standard items. These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

In this paper we describe an automated method of classifying research project descriptions: a human expert classifies a sample set of projects into a set of disjoint and pre-defined classes, and then the computer learns from this sample how to classify new projects into these classes. Both textual and non-textual information associated with the projects are used in the learning and classification phases. Textual information is processed by two methods of analysis: a natural language analysis followed by a statistical analysis. Non-textual information is processed by a symbolic learning technique. We present the results of some experiments done on real data: two different classifications of our research projects.

Uncertainty in the contents of a data base can be due to several reasons: errors in the data which is entered, changes in the real data which have not been introduced into the data base in the form of updates, errors in the data collection process, unreliable operation of the computer system, "don't care" conditions which are purposely left open by the data base designer, etc. The purpose of this paper is to present a formal model of uncertainty in terms of a probabilistic representation of the data base, and to evaluate the effect of this uncertainty on query processing and on the aggregate or summary information which may suffice in many applications. Our model leads to precise quantifiable engineering estimates and to theorems on the robustness of answers to queries as a function of the uncertainty in the data.