Memory scanning: Are fixed and varied sets scanned concurrently or successively?

Subjects were timed as they decided whether individually presented probe letters were or were not contained in pairs of memorized sets of one, two, or four letters. One set was fixed in advance of a block of trials, while the other varied from trial to trial. In Experiments 1 and 2, subjects responded positively if the probe was in either set and negatively otherwise; in Experiment 3 they made different responses to fixed-set, varied-set, and negative probes. In all three experiments, reaction time (RT) to varied-set probes depended more on the size of the varied set than of the fixed set, whereas RT to fixed-set and negative probes depended more on the size of the fixed set. These results were generally consistent with the hypothesis that scanning of the two sets was at least to some extent concurrent rather than successive, with the fixed set scanned more slowly than the varied set. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Efficient Management of Persistent Knowledge

Although computer speed has steadily increased and memory is getting cheaper, the need for storage managers to deal efficiently with applications that cannot be held into main memory is vital. Dealing with large quantities of clauses implies the use of persistent knowledge and thus, indexing methods are essential to access efficiently the subset of clauses relevant to answering a query. We introduce PerKMan, a storage manager that uses G-trees, and aims at efficient manipulation of large amounts of persistent knowledge. PerKMan may be connected to Prolog systems that offer an external C language interface. As well as the fact that the storage manager allows different arguments of a predicate to share a common index dimension in a novel manner, it indexes rules and facts in the same manner. PerKMan handles compound terms efficiently and its data structures adapt their shape to large dynamic volumes of clauses, no matter what the distribution. The storage manager achieves fast clause retrieval and reasonable use of disk space.     

CALCE/RAMCAD for Electronics

The University of Maryland's Computer Aided Life Cycle Engineering (CALCE) research effort is aimed at developing techniques for design, analysis, prediction, and optimization of electronics with respect to a variety of design goals. Performance, producibility, reliability, maintainability, life-cycle cost, safety, etc., are the design criteria which are being integrated through the application of interactive heuristic and algorithmic design optimization techniques. The University of Maryland Reliability and Maintainability Computer-Aided Design (RAMCAD) research is CALCE and is focused on second- and third-level electronic package design with respect to interfacing thermal, mechanical, and reliability analysis. The goal is to make reliability and maintainability an integral, real-time, and upfront part of the design process, such that the need to resubmit designs will be eliminated. The motivation behind this project was the fact that schedule constraints are often compromising the effectiveness and efficiency of the design process. To achieve the goal of CALCE/RAMCAD, a cohesive and extensible design environment for analysis and tradeoff tools was developed. A system executive controller that integrates tools for analysis while providing system transparency, design data management, and process management was developed to guide the user through the design process and provide direct access to the design tools. These developments required the integration of computer-aided design (CAD), artificial intelligence (AI), and decision support (DS) techniques. Attention to human factors and the use of interactive graphics were also instrumental in the development.     

The “burning ship” and its quasi-Julia sets

By manipulating the imaginary part of the complex-analytic quadratic equation, we obtain, by iteration, a set that we call the “burning ship” because of its appearance. It is an analogue to the Mandelbrot set (M-set). Its nonanalytic “quasi”-Julia sets yield surprizing graphical shapes.     

Effects of design parameters on geared two-link mechanisms

The design of a specialized agricultural harvesting machine required the use of a path generation mechanism that was compact, accurate, inexpensive to manufacture and difficult attempts involved the use of a cam and follower, but they were expensive to manufacture and difficult to operate. This paper discusses an alternative method of using a coupled, open loop, two-link mechanism to generate a polygon-like closed path (pseudo-polygon) with an integer number of curved sides and apexes. A complete development of the path and path curvaturei equations are presented. The influence of the design variables on the coupler shape (flatness of each side) is shown in equation and design chart form. An example of a successful application of this design method applied to a plant cutting mechanism is presented.     

Learning and Revising User Profiles: The Identification of Interesting Web Sites

We discuss algorithms for learning and revising user profiles that can determine which World Wide Web sites on a given topic would be interesting to a user. We describe the use of a naive Bayesian classifier for this task, and demonstrate that it can incrementally learn profiles from user feedback on the interestingness of Web sites. Furthermore, the Bayesian classifier may easily be extended to revise user provided profiles. In an experimental evaluation we compare the Bayesian classifier to computationally more intensive alternatives, and show that it performs at least as well as these approaches throughout a range of different domains. In addition, we empirically analyze the effects of providing the classifier with background knowledge in form of user defined profiles and examine the use of lexical knowledge for feature selection. We find that both approaches can substantially increase the prediction accuracy.     

Energy pricing policies in developing countries

Government energy pricing policies have multiple and often conflicting objectives: economic efficiency, government revenues (for both parastatal supply companies and the treasury), maintenance or improvement of income distribution, promotion of particular sectors (such as industrial exporters and local resource development) demand management and security of supply. It is important to examine the impacts on and the trade-offs between these objectives resulting from alternative policies in order to assist in policy selection. This article discusses the more important objectives and their conflicts and outlines an approach for the quantitative examination of alternative policies.     

Acidity in upland and hill soils: Cation exchange capacity,pH and lime requirement

For 30 acid upland and hill soils, with initial pH from 3.5 to 5.8, and loss on ignition from 14% to 94%, cation exchange capacity varied with pH according to the relationship CEC = m pH+C, where m and C are characteristic of each soil. The slope m depended on loss on ignition (LOI), and is given by: m=Δ CEC/Δ=0.22 LOI+1.92. Because of these relationships, CEC measured at the natural pH of the soil (in 0.025 M MgSO₄ solution) is more relevant to field conditions than a conventional measurement at pH7. An improved method for estimating lime requirement (LR) of these soils was developed, based on the relationship found between LR and the difference in CEC (ΔCEC) between that at the natural pH of the soil and that at the target pH of 5.5: LR=1.12 ΔCEC+1.48. Lime requirement can therefore also be estimated from loss on ignition using the earlier relationships, and is given by: LR=(0.25 LOI+2.15) Δ+1.48, where Δ is the difference between the natural pH (in 0.025 M MgSO₄) and the target pH. Predictions based on this agreed well with results for a further 10 soils.     

Sensitization of the startle reflex by footshock: Blockade by lesions of the central nucleus of the amygdala or its efferent pathway to the brainstem.

Bilateral electrolytic lesions of the central, but not the lateral, nucleus of the amgydala blocked shock sensitization of startle (the increase in startle produced by presentation of ten 0.6-mA footshocks in rapid succession). Lesions of the central nucleus also decreased reactivity to shock (jumping and flinching) during shock presentation. However, this decrease in reactivity cannot account for the blockade of shock sensitization, because when a higher shock intensity (1.0 mA) was used, producing equivalent reactivity to that of controls at 0.6 mA, central nucleus lesions still blocked shock sensitization. Moreover, lesions of the caudal part of the ventral amygdalofugal pathway, which carries central nucleus efferents to the startle reflex pathway, also blocked shock sensitization. It is hypothesized that shock activates the central nucleus of the amygdala, which increases startle through modulation of the startle pathway. Activation of the amygdala by shock may be the unconditioned response relevant for fear conditioning. (PsycINFO Database Record (c) 2010 APA, all rights reserved)