Abstract
Interest in the MapReduce programming model has been rekindled by Google in the past 10 years; its popularity is mostly due to the convenient abstraction for parallelization details this framework provides. State-of-the-art systems such as Google's, Hadoop or SAGA often provide added features like a distributed file system, fault tolerance mechanisms, data redundancy and portability to the basic MapReduce framework. However, these features pose an additional overhead in terms of system performance. In this work, we present a MapReduce design for Prolog which can potentially take advantage of hybrid parallel environments; this combination allies the easy declarative syntax of logic programming with its suitability to represent and handle multi-relational data due to its first order logic basis. MapReduce for Prolog addresses efficiency issues by performing load balancing on data with different granularity and allowing for parallelization in shared memory, as well as across machines. In an era where multicore processors have become common, taking advantage of a cluster's full capabilities requires the hybrid use of parallelism.

Abstract
Tabling is a technique that overcomes some limitations of traditional Prolog systems in dealing with redundant sub-computations and recursion. When tabling is combined with multithreading, we have the best of both worlds, since we can exploit the combination of higher declarative semantics with higher procedural control. To support this combination, the Yap Prolog system has, at engine level, multiple designs that vary from a No-Sharing design, where each thread allocates fully private tables, to a Full-Sharing (FS) design, where threads share the complete table space. In this work, we propose an extension to the table space data structures, which we named Private Answer Chaining (PAC), as way to support batched scheduling evaluation with the FS design. Batched scheduling is one of the most successful tabling scheduling strategies, known to be useful when a tabled logic program requires an eager propagation of answers and/or do not requires the complete set of answers to be found. Experimental results show that PAC is a good first approach, since with it the FS design remains quite competitive.

Abstract
Nowadays, clusters of multicores are becoming the norm and, although, many or-parallel Prolog systems have been developed in the past, to the best of our knowledge, none of them was specially designed to explore the combination of shared and distributed memory architectures. In recent work, we have proposed a novel computational model specially designed for such combination which introduces a layered model with two scheduling levels, one for workers sharing memory resources, which we named a team of workers, and another for teams of workers (not sharing memory resources). In this work, we present a first implementation of such model and for that we revive and extend the YapOr system to exploit or-parallelism between teams of workers. We also propose a new set of built-in predicates that constitute the syntax to interact with an or-parallel engine in our platform. Experimental results show that our implementation is able to increase speedups as we increase the number of workers per team, thus taking advantage of the maximum number of cores in a machine, and to increase speedups as we increase the number of teams, thus taking advantage of adding more computer nodes to a cluster.

Abstract
Map-Reduce is a programming model that has its roots in early functional programming. In addition to producing short and elegant code for problems involving lists or collections, this model has proven very useful for large-scale highly parallel data processing. In this work, we present the design and implementation of a high-level parallel construct that makes the Map-Reduce programming model available for Prolog programmers. To the best of our knowledge, there is no Map-Reduce framework native to Prolog, and so the aim of this work is to offer data processing features from which several applications can greatly benefit; the Inductive Logic Programming field, for instance, can take advantage of a Map-Reduce predicate when proving newly created rules against sets of examples. Our Map-Reduce model was comprehensively tested with different applications. Our experiments, using the Yap Prolog system, show that: (i) the model scales linearly up to 24 processors; (ii) a dynamic distributed scheduling strategy performs better than centralized or static scheduling strategies; and (iii) the performance varies significantly with the number of items being sent to each processor at a time. Overall, our Map-Reduce framework presents as a good alternative for both taking advantage of the currently available low cost multi-core architectures and developing scalable data processing applications, native to the Prolog programming language. © 2013 ACM.

Abstract
Logic Programming languages, such as Prolog, provide a high-level, declarative approach to programming. Despite the power, flexibility and good performance that Prolog systems have achieved, some deficiencies in Prolog's evaluation strategy - SLD resolution - limit the potential of the logic programming paradigm. Tabled evaluation is a recognized and powerful technique that overcomes SLD's susceptibility in dealing with recursion and redundant sub-computations. In a tabled evaluation, there are several points where we may have to choose between different tabling operations. The decision on which operation to perform is determined by the scheduling algorithm. The two most successful tabling scheduling algorithms are local scheduling and batched scheduling. In previous work, we have developed a framework, on top of the Yap Prolog system, that supports the combination of different linear tabling strategies for local scheduling. In this work, we propose the extension of our framework to support batched scheduling. In particular, we are interested in the two most successful linear tabling strategies, the DRA and DRE strategies. To the best of our knowledge, no other Prolog system supports both strategies simultaneously for batched scheduling. Our experimental results show that the combination of the DRA and DRE strategies can effectively reduce the execution time for batched evaluation.

Abstract