Abstract
Nowadays applications produce infinite streams of data distributed across wide sensor networks. In this work we study the problem of continuously maintain a cluster structure over the data points generated by the entire network. Usual techniques operate by forwarding and concentrating the entire data in a central server, processing it as a multivariate stream. In this paper, we propose DGClust, a new distributed algorithm which reduces both the dimensionality and the communication burdens, by allowing each local sensor to keep an online discretization of its data stream, which operates with constant update time and (almost) fixed space. Each new data point triggers a cell in this univariate grid, reflecting the current state of the data stream at the local site. Whenever a local site changes its state, it notifies the central server about the new state it is in. This way, at each point in time, the central site has the global multivariate state of the entire network. To avoid monitoring all possible states, which is exponential in the number of sensors, the central site keeps a small list of counters of the most frequent global states. Finally, a simple adaptive partitional clustering algorithm is applied to the frequent states central points in order to provide an anytime definition of the clusters centers. The approach is evaluated in the context of distributed sensor networks, presenting both empirical and theoretical evidence of its advantages.

Abstract
Logtalk, an object oriented logic programming language, provides experimental support for multi-threading programming with selected back-end Prolog compilers. By making use of core, low-level Prolog predicates that interface with operating-system native threads, Logtalk provides a high-level set of directives and predicates that allows programmers to easily take advantage of modern multi-processor and multi-core computers without worrying about the details of creating, synchronizing, or communicating with threads. Logtalk multi-threading programming features include support for concurrent calls akin to and-parallelism and or-parallelism, non- deterministic thread goals, asynchronous calls, and predicate synchronization. The integration with the Logtalk object-oriented features allows objects to send and receive both synchronous and asynchronous messages and to call local predicates concurrently. Logtalk multi-threading features are orthogonal to object-oriented concepts and can be useful even in the context of plain Prolog.

Abstract
This paper presents the logic programming concept of thread-based competitive or-parallelism,, which combines the original idea of competitive or-parallelism with committed-choice nondeterminism and speculative threading. In thread-based competitive or-parallelism, an explicit. disjunction of subgoals is interpreted as a set of concurrent, alternatives, each running in its own thread. The subgoals compete for providing an answer and the first successful subgoal leads to the nation of the remaining ones. We discuss the implementation of competitive or-parallelism in the context, of Logtalk, all object-oriented logic programming language, and present results.

Abstract
The amount of data collected and stored in databases is growing considerably in almost all areas of human activity. In complex applications the data involves several relations and proposionalization is not a suitable approach. Multi-Relational Data Mining algorithms can analyze data from multiple relations, with no need to transform the data into a single table, but are computationally more expensive. In this paper a novel relational classification algorithm based on the k-nearest neighbour algorithm is presented and evaluated.

Abstract
Tabled evaluation has been proved an effective method to improve several aspects of goal-oriented query evaluation, including termination and complexity. Several "native" implementations of tabled evaluation have been developed which offer good performance, but many of them require significant changes to the underlying Prolog implementation, including the compiler and the abstract machine. Approaches based on program transformation, which tend to minimize changes to both the Prolog compiler and the abstract machine, have also been proposed, but they often result in lower efficiency. We explore some techniques aimed at combining the best of these worlds, i.e., developing an extensible implementation which requires minimal modifications to the compiler and the abstract machine, and with reasonably good performance. Our preliminary experiments indicate promising results.

Abstract
Arguably. the most successful data, structure for tabling is tries. However, while tries are very efficient for variant based tabled evaluation, they are limited in their ability to recognize and represent, repeated terms in different tabled calls or/and answers. lit this paper, we propose a new design for the table space where tabled terms are stored in a common global trie instead of being spread over several different tries.