Transactions on Large-Scale Data- and Knowledge-Centered Systems XXII

An abundance of data generated from a multitude of sources, and intelligence derived by analyzing the same, has become an important asset across many walks of life. Simultaneously, it raises serious concerns about privacy. Differential privacy has become a popular way to reason about the amount of information about individual entries of a dataset that is divulged upon giving out a perturbed result for a query on a given data-set. However, current differentially-private algorithms are computationally inefficient, and do not explicitly exploit the abundance of data, thus wearing out the privacy budget irrespective of the volume of data. In this paper, we propose BPMiner, a solution that is both private and accurate, while simultaneously addressing the computation and budget challenges of very big datasets. The main idea is a non-trivial combination between differential privacy, sample-and-aggregation, and a classical statistical methodology called sequential estimation. Rigorous proof regarding the privacy and asymptotic accuracy of our solution are provided. Furthermore, experimental results over multiple datasets demonstrate that BPMiner outperforms current private algorithms in terms of computational and budget efficiencies, while achieving comparable accuracy. Overall, BPMiner is a practical solution based on strong theoretical foundations for privacy-preserving analysis on big datasets.

Nowadays, digital systems are connected through complex architectures. These systems involve persons, physical and digital resources such that we can consider that a system consists of elements from two worlds, the social world and the digital world, and their relations. Users perform activities like chatting, buying, sharing data, etc. Evaluating and choosing appropriate systems involve aspects like functionality, performance, QoS, ease of use, or price. Recently, trust appeared as another key factor for such an evaluation. In this context, we raise two issues, (i) how to formalize the entities that compose a system and their relations for a particular activity? and (ii) how to evaluate trust in a system for this activity? This work proposes answers to both questions. On the one hand, we propose SocioPath, a metamodel based on first order logic, that allows to model a system considering entities of the social and digital worlds and their relations. On the other hand, we propose two approaches to evaluate trust in systems, namely, SocioTrust and SubjectiveTrust. The former is based on probability theory to evaluate users’ trust in systems for a given activity. The latter is based on subjective logic to take into account uncertainty in trust values.

With the advent of Web 2.0 users are producing bigger and bigger amounts of diverse data, which are stored in a large variety of systems. Since the users’ data spaces are scattered among those independent systems, data sharing becomes a challenging problem. Distributed search and recommendation provides a general solution for data sharing and among its various alternatives, gossip-based approaches are particularly interesting as they provide scalability, dynamicity, autonomy and decentralized control. Generally, in these approaches each participant maintains a cluster of “relevant” users, which are later employed in query processing. However, as we show in the paper, only considering relevance in the construction of the cluster introduces a significant amount of redundancy among users, which in turn leads to reduced recall. Indeed, when a query is submitted, due to the high similarity among the users in a cluster, the probability of retrieving the same set of relevant items increases, thus limiting the amount of distinct results that can be obtained.In this paper, we propose a gossip-based search and recommendation approach that is based on diversity-based clustering scores. We present the resultant new gossip-based clustering algorithms and validate them through experimental evaluation over four real datasets, based on MovieLens-small, MovieLens, LastFM and Delicious. Compared with state of the art solutions, we show that taking into account diversity-based clustering score enables to obtain major gains in terms of recall while reducing the number of users involved during query processing.

The ever-growing literature in biomedicine makes it virtually impossible for individuals to grasp all the information relevant to their interests. Since even experts’ knowledge is limited, important associations among key biomedical concepts may remain unnoticed in the flood of information. Discovering those hidden associations is called hypothesis discovery or literature-based discovery. This paper propose an approach to this problem taking advantage of a closed, triangular chain of relations extracted from the existing literature. We consider such chains of relations as implicit rules to generate explanatory hypotheses. The hypotheses generated from the implicit rules are then compared with newer knowledge for assessing their validity and, if validated, they are served as positive examples for learning a regression model to rank hypotheses. As a proof of concept, the proposed framework is empirically evaluated on real-world knowledge extracted from the biomedical literature. The results demonstrate that the framework is able to produce legitimate hypotheses and that the proposed ranking approach is more effective than the previous work.

Given the vast area to be covered and the random deployment of the sensors, wireless sensor networks (WSNs) require scalable architecture and management strategies. In addition, sensors are usually powered by small batteries which are not always practical to recharge or replace. Hence, designing an efficient architecture and data management strategy for the sensor network are important to extend its lifetime. In this paper, we propose energy efficient two-level data aggregation technique based on clustering architecture with which data is sent periodically from nodes to their appropriate Cluster-Heads (CHs). The first level of data aggregation is applied at the node itself to eliminate redundancy from the collected raw data while the CH searches, at the second level, nodes that generate redundant data sets based on the variance study with three different Anova tests. Our proposed approach is validated via experiments on real sensor data and comparison with other existing data aggregation techniques.