Search

In data mining, searching for frequent patterns is a common basic operation. It forms the basis of many interesting decision support processes. In this paper we present a new type of patterns, binary expressions. Based on the properties of a specified binary test, such as reflexivity, transitivity and symmetry, we construct a generic algorithm that mines all frequent binary expressions. We present three applications of this new type of expressions: mining for rules, for horizontal decompositions, and in intensional database relations. Since the number of binary expressions can become exponentially large, we use data mining techniques to avoid exponential execution times. We present results of the algorithm that show an exponential gain in time due to a well chosen pruning technique.

The following study investigates and identifies research trajectories pertaining to the digital transition of agriculture and food production. While a vast amount of research aims to discover new technologies, or to apply them in novel ways, their large-scale implications as regards data ownership and data governance are relatively overlooked. Regulatory interventions are demanded to steer data ownership and data governance towards the ‘common good’. It is thus necessary to identify how research can contribute to the discussion on sensitive areas of policymaking that have been the object of environmental regulation, including the EU Green Deal and UN Sustainable Development Goals. In the light of this necessity, this paper identifies issues with ethical relevance emerging from the adoption of new technologies in agritech, including Artificial Intelligence techniques and Internet of Things applications. To do so, this study attempts to systematise and categorise existing research trends by clearly identifying their scope and understanding the relationships that exist among them. The results of this enquiry show that five interconnected research trajectories- namely, technical solutions, data governance, data ownership, ethics and law- can foster the discussion on agritech transition. The connections between these research areas can be understood in terms of a descriptive and a prescriptive perspective.

Two years after the General Data Protection Regulation (GDPR) went into effect, data anonymization remains one of the main issues linked to fragmen-tation in the Member States’ anonymization policies, in which regard stakehold-ers would like additional guidelines. In keeping with this premise, this article aims to analyze and compare trends in the implementation and enforcement of anonymization policies put in place by data protection authorities in two countries: France and Italy. This analysis makes it possible to trace the evolution of these policies and high-light their critical importance in applying the data minimization principle and in enforcing the right to erasure under Art. 17 GDPR.

Best Paper Award; International audience; Many scientific experiments are performed using scientific workflows, which are becoming more and more data-intensive. We consider the efficient execution of such workflows in the cloud, leveraging the heterogeneous resources available at multiple cloud sites (geo-distributed data centers). Since it is common for workflow users to reuse code or data from other workflows, a promising approach for efficient workflow execution is to cache intermediate data in order to avoid re-executing entire workflows. In this paper, we propose a solution for distributed caching of scientific workflows in a multisite cloud. We implemented our solution in the OpenAlea workflow system, together with cache-aware distributed scheduling algorithms. Our experimental evaluation on a three-site cloud with a data-intensive application in plant phenotyping shows that our solution can yield major performance gains, reducing total time up to 42% with 60% of same input data for each new execution.

An audio classifier that can distinguish between speech, music, silence and garbage has been developed. The classifier was trained and tested on broadcast news material provided by VRT (Flemish Radio and Television Network). Several feature sets and machine learning algorithms have been tested, providing choices of speed and performance for a target system. The audio classifier is part of a greater system that together with visual data can retrieve information from news broadcasts: speech can be converted to text and the speaker can be recognized. Music can be further used for genre classification, jingle recognition or copyright infringement detection. Silence is recognized and used to provide cues on topic changes or speaker turns. At this point everything that is not classified as speech, music or silence is labeled garbage. Garbage classes can be further used for background categorization giving information on the environment where someone speaks (an anchor in the studio or a reporter in the street)