Knowledge elicitation using more than one expert to cover the same domain

Knowledge elicitation is accepted as being one of the most problematic areas in the creation of a knowledge-based system.A large amount of research has already concentrated on finding more efficient and effective techniques for eliciting knowledge from an individual expert. However, little attention has been given to the involvement of more than one source of expertise in knowledge-based system development.This paper is based on the authors' practical experience gained when developing a knowledge-based system for the conceptual design of bridges. It shows that the use of more than one expert throughout the knowledge elicitation process can improve both the efficiency of the approach and the quality of the knowledge acquired.     

Optical circulator consisting of a YIG spherical lens, PANDA-fibre polarisers and a fibre-optic polarising beam splitter

A novel optical circulator is reported which employs an yttrium-iron-garnet (YIG) sphere, PANDA-fibre polarisers and a fibre-optic polarising beam splitter. The splitting ratios of the fibre-optic polarising beam splitter were 1.0:99.0 for x-polarisation and 97.5:2.5 for y-polarisation, with an excess loss of 0.5 dB at 1.30 ?m wavelength. The backward isolation of the optical circulator was 30 dB with the total insertion loss of 4.5 dB.     

An improved scan laser with a VO2programmable mirror

A 10.6 μm scan laser has been constructed and operated with an off-axis cathode ray tube, high reflectance multilayer thin-film structures, and a tapered plasma discharge tube. Equations are given for the switching time of a high-reflectance spot on the VO2and for the relation of scan laser output power to cavity geometry, cavity losses, and the gain of the active CO2medium. A scan capability of2.1 times 10^{3}easily resolvable directions was demonstrated, and sequential and randomly addressed spot rates of 10⁵/s were achieved. The equations relating output power and cavity mode size were experimentally verified using a nonscanned beam.     

The effect of plastic deformation on Al2CuLi (T 1) precipitation

The enhancement ofT ₁ precipitation in Al-Li-Cu alloys by plastic deformation prior to aging (that is, cold work) and the subsequent increase in alloy strength is investigated. The increased understanding of the role of matrix dislocations in the nucleation and growth ofT ₁ plates, discussed in the previous paper,^[1] permits a detailed study of the phenomenon. In this paper, the effect of different levels of plastic strain on theT ₁ particle distributions as a function of aging time at 190 °C is quantified, and the subsequent influence on tensile properties is thereby described. The effect of plastic deformation is shown to decrease theT ₁ plate length and thickness, increase the number density by almost two orders of magnitude, increase the yield strength by 100 MPa, while simultaneously reaching peak strength in 20 pct of the time required without plastic deformation. Formerly Graduate Student, Department of Materials Science, University of Virginia,     

Sensory analysis of polystyrene packaging material taint in cocoa powder for drinks and chocolate flakes

Polystyrene packaging material taint was sensorily evaluated in cocoa powder for drinks and chocolate flakes using short-cut signal detection measures on differences between control and test samples and on recognition of styrene. No differences were observed in cocoa powder for drinks and plain chocolate flakes treated with 0.5 dm2 polystyrene of 1 mm thickness. However, differences were detected in milk chocolate flakes and plain chocolate flakes, which were in contact with a larger area or thicker polystyrene packaging material. The latter results were confirmed by the styrene recognition test, so polystyrene is a potential source of off-flavour for chocolate products. The amount of residual styrene in the polystyrene used was about 320 ppm, while the amounts of styrene ranged from 7 to 132 ppb in cocoa drinks and from 414 to 1447 ppb in chocolate flakes.     

Construction and performance of a multiparameter focal plane detector for use on a QMG/2 magnetic spectrometer

The design, construction and performance of a modular, two pressure focal plane detector, 550 mm long, 1000 mm deep and with an active height of 60 mm are described. The detector has been operated with ions ranging from 50 MeV ¹H to 200 MeV ³⁶S. Typical results achieved are 0.5 mm and 0.3° for the spatial and angular resolutions, and 0.7% and 3% for the total energy and energy loss measurements. A height signal is also provided.     

Electrochemical Faceting of polycrystalline platinum in HF solutions

The electrochemical faceting of polycrystalline (pc) platinum yielding a (100)-preferred oriented electrode surface was accomplished in HF solutions by applying to the pc platinum electrode a square wave potential perturbation at 1 kHz in the 0.2 to 1.3 V range (vs rhe). SEM patterns of the resulting surfaces were also obtained. For HF concentrations up to 2 M no appreciable influence of the electrolyte concentration on the electrochemical faceting was noticed. Otherwise, the voltammetric response of the (100)-type preferred oriented platinum surface in the H-adatom potential range when compared to previously reported data resulted remarkably sensitive to the electrolyte composition. The discussion of results was based on the non-equilibrium adsorption conditions for anions under the periodic perturbation.     

Location errors in tachistoscopic recognition: Guesses, probe errors, or spatial confusions?

Examined whether location errors in tachistoscopic visual probe tasks reflect (1) uncertainty about the actual position of items, (2) inappropriate guesses, encouraged by a requirement to respond, or (3) confusion about the position being tested. 46 undergraduates completed probe tasks in which they provided confidence ratings on each trial and tasks that substituted a color cue for a visual pointer or used a double bar probe. Results suggest that location errors on visual probe tasks reflect confusion about the spatial position of target items. Findings are consistent with data questioning the concept of iconic memory and imply that the bar-probe task cannot be used to study precategorical visual memory. (French abstract) (PsycINFO Database Record (c) 2010 APA, all rights reserved)