Recently, several studies have investigated the efficient exe-cution of bag-of-words (BOW) retrieval models for large doc-ument collections. Even though there is a large body of work that consistently shows that dependency models consistently outperform BOW models, none of these studies consider

with quite reasonable efficiency. Execution is achieved by generating Common Lisp code from PVS. The key step in generating efficient code is the use of static analysis to determine safe destructive array updates. Contents 1 Introduction 5 2 A Brief Summary of the PVS Specification Language 7 3 Executing

techniques for the efficient execution of KPNs, taking into account both execution time and memory usage. Keywords: Kahn process networks, dynamic scheduling, deadlock resolution, multi-processor architectures, media processing, signal processing 1

Abstract not found

. It then uses a novel and powerful form of local completeness reasoning in order to transform those plans into contingent plans of high quality. These contingent plans can be efficiently executed, obtaining more answers at less cost than the original plans. We focus in this paper on the algorithms underlying

events are adjusted appropriately. Unfortunately, this may run afoul of tight real-time control requirements that dictate extreme efficiency. The performance may be improved by restricting the propagation. However, a fast, locally propagating, execution controller may incorrectly execute a consistent

contains the synchronous guarded actions. In order to execute the synchronous guarded actions efficiently, a simple vectorization algorithm is intro-duced and the limiting factors of the graphics device are emphasized. Additionally, a general perspective to the performance of optimized CUDA programs which

This paper proposes a method to restrict the rule appliction by activating rules dynamically. The logical relationship among rules are pre-a]alyzed nd a set of anteced]t actions, which are prerequisite for the condition of tJe rule being satisfied is determined for ,each ruleo In execution tin a

Memoization is an optimization that provides asymptotic speedups, automatically achieving many of the benefits of dynamic programming. Memoization, however, trades off reduced execution time for additional required storage. This additional storage requirement can be reduced somewhat by using

environment