FY 2000 ITL Publications

Note that some documents are published in more than one place. Due to the large number of documents, publications listed in previous ITL Technical Accomplishment reports are not repeated.

Ammann, P.E., Black, P.E.; A Specification-Based Coverage Metric to Evaluate Test Sets; NISTIR 6403 and 4th IEEE International; Symposium on High Assurance Systems Engineering (HASE 99), Washington, DC, November 1999. Paper extended and revised for International Journal of Reliability, Quality and Safety Engineering, December 2000; 10/31/1999

Software developers use a variety of methods, including both formal methods and testing, to argue that their systems are suitable components for high assurance applications. In this paper, we develop another connection between formal methods and testing by defining a specification-based coverage metric to evaluate test sets. Formal methods in the form of a model checker supply the necessary automation to make the metric practical. The metric gives the software developer assurance that a given test set is sufficiently sensitive to the structure of an application's specification. In this paper, we develop the necessary foundation for the metric and then illustrate the metric on an example.

Ammann, P.E., Black, P.E.; Abstracting Formal Specifications to Generate Software Tests via Model Checking; NISTIR 6405 and 18th Digital Avionics Systems Conference, St. Louis, Missouri; 10/31/1999

A recent method combines model checkers with specification-based mutation analysis to generate test cases from formal software specifications. However, high-level software specifications usually must be reduced to make analysis with a model checker feasible. We propose a new reduction, parts of which can be applied mechanically, to soundly reduce some large, even infinite, state machines to manageable pieces. Our work differs from other work in that we use the reduction for generating test sets as opposed to the typical goal of analyzing for properties. Consequently, we have different criteria, and we prove a different soundness rule. Informally the rule is that counterexamples from the model checker are test cases for the original specification. The reduction changes the state machine and temporal logic constraints in the model checking specification to avoid generating unsound test cases. We give an example of the reduction and test generation.

Anderson, D.M., Boettinger, W.J., McFadden, G.B., Wheeler, A.A.; A Phase-Field/Fluid Motion Model of Solidification: Investigation of Flow Effects During Directional Solidification and Dendritic Growth; Proceedings of the NASA Microgravity Materials Science Conference, Huntsville, Alabama, June 6-8, 2000; 7/6/2000

The phase-field model of diffusion-controlled solidification has recently been extended to include the effects of fluid flow in the melt. The phase-field model is based on coupling the equations for heat flow in the liquid and solid phases with an auxiliary equation that describes the evolution of the phase-field variable, which is a non-conserved order parameter indicating the local phase, solid or liquid, at each point of the material. The solid-liquid interface is then represented by a diffuse transition layer in which the phase-field variable changes rapidly between its values in the bulk phases. The extended model includes fluid flow by a further coupling to the Navier-Stokes equations. In our work, the solid phase is treated as a fluid of high viscosity compared to the liquid phase. The main coupling in the Navier-Stokes equations is then through an additional term in the stress tensor that depends on the gradients of the phase-field variable, representing the effects of capillary forces within the diffuse interface. This model is applied to solidification and crystal growth situations in order to investigate the effect of fluid motion in the melt on the growth characteristics.

Anderson, D.M., McFadden, G.B., Wheeler, A.A.; A Phase-Field Model with Convection: Numerical Simulations; NISTIR 6442 and to appear in Proceedings on Interfaces for the Twenty-First Century, Imperial College Press, London, England, 2000

In a previously developed phase-field model of solidification that includes convection in the melt, the two phases are represented as viscous liquids, where the putative solid phase has a viscosity much larger than the liquid phase. In this paper, we report numerical computations on a simplified form of this model which represents the growth of a two-dimensional dendrite in a thin gap between two parallel thermally insulting plates. In these computations, flow in the liquid arises because of the differing densities of the solid and liquid phases.

Auzanne, C.G.P., Garofolo, J.S., Fiscus, J.G., Fisher, W.M.; Automatic Language Model Adaptation for Spoken Document Retrieval; RIAO-2000 Content-Based Multimedia Information Access Conference, Paris, France, 2000

This paper describes experiments implemented at NIST in adapting language models over time to improve recognition of broadcast news recorded over many months. These experiments were designed specifically to improve the utility of automatically generated transcripts for retrieval applications. To evaluate the potential of the approach, a time-adaptive automatic speech recognition run was implemented to support the 1999 TREC Spoken Document Retrieval (SDR) Track - more than 500 hours of broadcast news sampled across five months. The accuracy of retrieval for several systems using the time-adaptive system transcripts was evaluated against transcripts produced by virtually the same recognition system with a fixed language model. This paper details the process we employed to identify and implement the time-adaptive language model and discusses the results of the experiment in terms of its effect on word error rate, out of vocabulary rate and retrieval accuracy (Mean Average Precision).

Babb, D.M., Verlinde, J., Rust, B.W.; The Removal of Turbulent Broadening in Radar Doppler Spectra Using Linear Inversion With Double-Sided Constraints; Journal of Atmospheric and Oceanic Technology

Remote sensing instruments have the ability to collect data over extensive temporal periods and spatial regions. A common thread between all these sensors is the need to relate the measured quantity to a meaningful observation of a system property. If the relationship between each measurement and the set of atmospheric quantities that influence that measurement is known, the problem can be reduced to a set of linear equations. Solving for the unknown atmospheric quantities then becomes a linear algebra problem where the solution vector is equal to the inverse of the kernel matrix multiplied by the set of independent measurements. However, in most remote sensing applications, inversion of the kernel matrix is unstable resulting in the amplification of measurement and computational uncertainties. Techniques to circumvent this error amplification have focused on methods of constraining the solution. In this paper, we adapt an existing technique to do such an inversion. Noise reduction is accomplished by the addition of double-sided inequality constraints for each unknown variable. The advantage of such a technique is the ability to individually adjust the solution space of each individual unknown depending on a prior knowledge.

The inversion algorithm is applied to the problem of retrieving radar Doppler spectra, which have been artificially broadened by turbulent air motions. First, to test the algorithm, radar Doppler spectra were simulated using known drop-size and vertical air motion distributions. The simulated spectra were used as input to the retrieval algorithm and the results compared to the initial quiet-air spectrum. Results indicate that accurate retrievals can be performed despite the addition of moderate amounts of noise to the simulated spectra. Then, to demonstrate the practical retrieval of quiet-air Doppler spectra, the algorithm was used to process radar observations collected from continental stratocumulus. From these retrievals, a 2-D map of the large-scale vertical motions within the cloud was constructed as well as a profile of vertical velocity variance. In addition, a drop-size distribution was also derived from an updraft region of the cloud.

Bassham III, Lawrence E.; Efficiency Testing of ANSI C Implementations of Round 2 Candidate Algorithms for the Advanced Encryption Standard; Proceedings of the Third Advanced Encryption Standard (AES) Candidate Conference, April 13-14, 2000; 4/13/2000

This paper describes the testing methodology used in ANSI C efficiency testing of the Round 2 Candidates for the Advanced Encryption Standard, along with a few observations regarding those measurements. The paper also presents tables of timing and cycle counting values obtained from testing the algorithms. Finally, there are some conclusions regarding which algorithms have the best (i.e., most consistent) performance.

Beichl, I., Sullivan, F.; The Metropolis Algorithm; IEEE Computing in Science and Engineering Magazine, 2, pp. 65-69, January/February 2000; 1/1/2000

This is a survey of the Metropolis Algorithm for the IEEE special issue on the ten most important algorithms of the century.

Beichl, I., Sullivan, F.; A = B; Accepted by IEEE Computing in Science and Engineering

This is a tutorial article for the "Computing Prescriptions" column. It explains how to use modular arithmetic in a large computation and why it would be useful.

Beichl, I., Sullivan, F.; Determining the Determinant; IEEE Computing in Science and Engineering

This tutorial article explains the Bareiss algorithm for finding the determinant of an integer matrix using exact arithmetic. It is then combined with the Chinese Remainder Theorem to give a method for matrices with extremely large entries.

Black, Paul E.; Is "Implementation Implies Specification" Enough?; Proceedings of 36th Design Automation Conference (DAC 2000), Los Angeles, California, June 5-9, 2000; 6/5/2000

An implementation is typically checked against a specification by proving that the implementation implies the specification. This ensures that the implementation only has behaviors allowed by the specification. However, this does not require the implementation to have any behavior to all! We propose that correctness statements have two parts, corresponding to liveness and safety. Safety is that the implementation implies an "allowed behavior" specification, as now. Liveness is that a "required behavior" specification implies the implementation.

Boisvert, Ronald F.; Mathematical Software: Past, Present, and Future; To appear in Proceedings of the International Symposium on Computational Science and Mathematics and Computers in Simulation

This paper provides some reflections on the field of mathematical software on the occasion of John Rice's 65th birthday. We describe some of the common themes of research in this field and recall some significant events in its evolution. Finally, we raise a number of issues that are of concern to future developments.

Brady, M.C., Montanez-Rivera, C., Rivello, R.M., Carnahan, L.; XML Technologies; ITL Bulletin, September 2000; 9/29/2000

An Internet language called the Extensible Markup Language (XML) is rapidly becoming one of the most popular languages in the world. XML is being incorporated into many Internet Web pages and applications; it is particularly useful for those involving structured information exchanges, such as electronic commerce. It is a language that describes information in a way that allows computers to exchange information and automatically act on the information. Consequently, it can speed up automation of certain processes. Surrounding XML is a family of technologies that either augment the language or further build upon it. It is sometimes difficult to sift through the various XML technologies and determine which ones best fit the needs of an organization. This ITL Bulletin introduces the key technologies needed to incorporate XML in your architecture, discusses programmatic methods for manipulating and displaying XML data, and identifies NIST resources that can be used in evaluating viable XML-related solutions.

Braun, R.J., Zhang, J., Cahn, J.W., McFadden, G.B., Wheeler, A.A.; Model Phase Diagrams for an FCC Alloy; NISTIR 6463 and Proceedings on Interfaces for the 21st Century, Imperial College Press, London, England, 2000

We describe a free energy model that allows the computation of phase diagrams for a particular set of solid-state phase transitions in a binary alloy. This is an extension of our previous work, in which we developed a diffuse-interface model of phase boundaries that occur in order-disorder transitions of a face-centered-cubic (FCC) binary alloy. That model was based on a fourth-order Landau expansion of the free energy function in terms of the three nonconserved order parameters that describe ordering on the underlying FCC lattice. In this paper, we examine an improved description of the free energy of the system that incorporates a more realistic dependence on the temperature and composition of the system. The improved model allows the computation of phase diagrams that delineate the states of equilibrium between the various phases that are described by the model. In particular, the phase transition between the disordered FCC phase and the ordered L10 phase is correctly described as a first order phase transition using the improved model, whereas in the previous model, this transition was instead modeled as second order. The improved model also allows the self-consistent calculation of concentration variations through interphase boundaries that separate bulk phases and through antiphase boundaries that separate variants of ordered phases.

Buckley, C., Voorhees, E.M.; Evaluating Evaluation Measure Stability; Proceedings of ACM SIGIR 2000 Conference; 7/24/2000

This paper presents a novel way of examining the accuracy of the evaluation measures commonly used in information retrieval experiments. It validates several of the rules-of-thumb experimenters use, such as the number of queries needed for a good experiment is at least 25 and 50 is better, while challenging other beliefs, such as the common evaluation measures are equally reliable. As an example, we show that Precision at 30 documents has about twice the average error rate as Average Precision has. These results can help information retrieval researchers design experiments that provide a desired level of confidence in their results. In particular, we suggest researchers using Web measures such as Precision at 10 documents will need to use many more than 50 queries or will have to require two methods to have a very large difference in evaluation scores before concluding that the two methods are actually different.

Burns, T.J., Davies, M.A.; On Repeated Adiabatic Shear Band Formation During High-Speed Machining; Submitted to International Journal of Plasticity

We compare the repeated adiabatic shear band formation that takes place at sufficiently large cutting speeds in a number of materials during high-speed machining operations with the more well-known formation of a single shear band that often takes place at sufficiently large strain rates in dynamic torsion tests on these materials. We show that there are several major differences in the physics of the two deformation processes. In particular, the shear stress in machining over the tool-material contact length is not even approximately homogeneous. Additionally, in high-speed machining, the material flow can become convection-dominated, so that the tool can "outrun" the thermal front generated in the work piece material by the high-strain-rate cutting process. We demonstrate by means of a simple one-dimensional continuum model that these differences can lead to repeated oscillations in the plastic flow of the work piece material during high-speed machining, leading to the repeated formation of adiabatic shear bands.

Carasso, Alfred S.; Direct Blind Deconvolution; NISTIR 6428 and accepted by SIAM Journal on Applied Mathematics; 11/17/1999

Blind deconvolution seeks to deblur an image without knowing the cause of the blur. Iterative methods are commonly applied to that problem, but the iterative process is slow, uncertain, and often ill-behaved. This paper considers a significant but limited class of blurs that can be expressed as convolutions of 2-D symmetric L�vy `stable' probability density functions. This class includes and generalizes Gaussian and Lorentzian distributions. For such blurs, a method is developed that can detect the point spread function from 1-D Fourier analysis of the blurred image. A separate image deblurring technique uses this detected point spread function to deblur the image. Each of these two steps uses direct non-iterative methods and requires interactive adjustment of parameters. Using this method, blind deblurring of 512 x 512 images can be accomplished in minutes of CPU time on current desktop workstations. Numerous blind experiments on synthetic data show that for a given blurred image, several distinct point spread functions may be detected that lead to useful yet visually distinct reconstructions.

Carnahan, Lisa J., Ruark, Marcus, Editors; Requirements For Real-time Extensions to the JavaT Platform: Report from the Requirements Group for Real-time Extensions for the JavaT Platform; NIST SP 500-243, http://www.nist.gov/rt-java; 3/8/2000

The National Institute of Standards and Technology (NIST) sponsored the Requirements Group for Real-time Extensions to the JavaT Platform. The Requirements Group includes representatives from companies and organizations whose expertise spans the computing industry and academia. Industry participants include desktop, server, and enterprise systems providers, embedded systems providers, device manufacturers, and real-time operating system vendors. The Requirements Group met during a six-month period in a series of open workshops. Additionally, the Requirements Group continued discussions on the rt-java@raleigh.ibm.com mailing list. The goal of the group was to develop cross-disciplinary requirements for real-time functionality that is expected to be needed by real-time applications written in the JavaT programming language and executing on various platforms. This document is the result of the Requirements Group's efforts.

Chandramouli, R.; Business Process Driven Framework for Defining an Access Control Service Based on Roles and Rules; Proceedings of the 23rd National Information Systems Security Conference, Baltimore, Maryland, October 16-19, 2000; 10/19/2000

Defining an Access Control Service for an enterprise application requires the choice of an access control model and a process for formulation of access decision rules to be used by the access enforcement mechanism. In this paper, we describe a business process driven framework (called the BPD-ACS) for developing both the model and formulating the access decision rules. The model used is the Role Based Access Control (RBAC) model and the access decision rules are based on temporal business associations. The enterprise setting is a multi-facility hospital and the particular application for which the access control service was defined is the Hospital-based Laboratory Information System. (HLIS).The lesson learned from this exercise is that a much more sophisticated rule processing capability is required for these types of applications than is currently available in both commercial and research-prototype authorization servers.

Chandramouli, R.; Application of XML Tools for Enterprise-Wide RBAC Implementation Tasks Fifth ACM Workshop on Role Based Access Control, Berlin, Germany, July 26-28, 2000; 7/26/2000

The use of XML and its associated APIs for information modeling and information interchange applications is being actively explored by the research community. In this paper, we develop a XML Document Type Definition (DTD) for representing the schema of a RBAC Model and a conforming XML document containing the actual RBAC-based access control data for a commercial banking application. Based on this DTD, the XML document and the methods in the DOM API Level 1.0 standards, we describe three application tasks related to enterprise-wide implementation of RBAC. They are (a) Implementing a RBAC model for a database application, (b) Implementing RBAC models with identical data on two different database servers, and (c) Transforming data under a RBAC model to a different but structurally similar model like the Group-based Access Control model. Other potential Access Control Service applications exploiting the capabilities of some commercial XML processors are also outlined.

Chang, Wo; Web-Based Annotation Collaboration Tool; NIST Journal of Research and International Journal of Software Engineering and Knowledge Engineering, February 2001

The Web has emerged as a dominant mode of making information available and is increasingly seen as the medium to support collaboration across boundaries of geography and time. To realize this potential, the Web must support viewing and manipulating of multi-source, multimedia objects that are interdependent, yet must be synchronized temporally and spatially. This paper presents a prototype tool, ACTS, which utilizes the W3C standard Web content technology, Synchronized Multimedia Integrated Language (SMIL), to provide a framework of temporal and spatial synchronization. ACTS is a National Institute of Standards and Technology, Information Technology Laboratory (NIST/ITL) prototype which is based on our S2M2 Java applet-based SMIL player that implements and extends the SMIL 1.0 specification. The ACTS prototype provides evidence that the Web can become an effective medium for collaboration over multimedia objects.

Choi, Jin Seek, Golmie, Nada, Su, David H.; A Bandwidth Guaranteed Multi-Access Protocol for WDM Local Networks; Proceedings of 17th International Conference on Communications, New Orleans, Louisiana, June 19-23, 2000; 6/19/2000

In this paper, we propose a bandwidth guaranteed multi-access protocol for broadcast-and-select WDM local networks with a star topology. The proposed protocol is based on a combination of contention and reservation mechanisms for time-slotted WDM networks. Every node accesses the data channel by transmitting request packets in minislots on a separate control channel. There are two types of minislots: reservation minislots and contention minislots. Nodes requiring bandwidth guarantees, called guaranteed nodes, use reservation minislots that are assigned by the control node. The remaining nodes share contention minislots using a random access mechanism. Each node dynamically assigns data channels for the minislots successfully returned on a First-Come-First-Served (FCFS) basis. Here, the number of reservation minislots is allocated by the control node. The remaining minislots are used for contention minislots. The reservation minislots can guarantee a minimum bandwidth for the guaranteed nodes. The contention minislots enable on-demand services at the optical layer and achieve good fairness for the remaining bandwidth. This protocol can be implemented with a simple distributed algorithm that efficiently utilizes the data channel.

Ciarletta, L.P., Dima, A.A.; A Conceptual Model for Pervasive Computing; Workshop on Pervasive Computing, the 29th International Conference on Parallel Computing 2000, Conference Proceedings, Toronto, Canada

As an emerging field, pervasive computing has not had the opportunity to develop a conceptual model similar to the OSI Reference Model used to describe computer networks. Such a model would be useful to properly classify issues raised during discussion and provide needed context. Inspired by the layers of abstraction provided by the OSI Reference Model, we present our Layered Pervasive Computing (LPC) model to facilitate discussion and analysis of pervasive computing systems by providing a much-needed conceptual framework. A key feature of our model is its representation of the human user at each layer of abstraction of the model. We will then use our model to analyze a research prototype created as part of our Aroma pervasive computing project. This analysis is illustrative because it quickly reveals issues that must be addressed to realize our research prototype as a commercial product.

Coakley, K.J.; Optimal Background Correction Strategies for Neutron Lifetime Experiments Using Magnetically Trapped Neutrons; Nuclear Instruments and Methods in Physics Research (A)

In the first stage of each run of a neutron lifetime experiment, a magnetic trap is filled with neutrons. In the second stage of each run, decay events plus background events are observed. In a separate experiment, background is measured. The mean lifetime is estimated by fitting a two parameter exponential model to the background corrected data. For two models of the background signal, I determine the optimal ratio of the number of "background only" measurements to the number of primary "neutron decay plus background" measurements. Further, for each run, I determine the optimal allocation of time for filling and for observing decay events. For the case where the background consists of activated aluminum plus a stationary Poisson process, the asymptotic standard error of the lifetime estimate computed from the background corrected data is lower than the asymptotic standard error computed from the uncorrected data. For the case where the background is a stationary Poisson process, background correction is desirable provided that the background intensity is sufficiently small compared to the rate at which neutrons enter the trap.

Coakley, K.J., Chowdhuri, Z., Richardson, J.M., Dewey, M.S.; Monte Carlo Modeling of Neutron Scattering Rocking Curve Data; Nuclear Instruments and Methods in Physics Research (A)

At NIST, an in-beam neutron lifetime experiment is underway. In part of the experiment, a neutron detector is calibrated. The accuracy of the detector calibration depends, in part, on how accurately the mean wavelength of a neutron beam can be estimated from rocking curve data. Here, rocking curve data is modeled using a Monte Carlo method. To speed up the simulation, an importance sampling method is used. Simulated and observed rocking curves are compared. For simulated data, the statistical bias of the mean wavelength estimate is found to be 0.004 percent.

Collins, D., Antonishek, J., Sell, S., Mink, A.; High-Speed Network Applications and Implications for Fiberoptic and Copper Connections; NISTIR 6536; 9/29/2000

We present an analysis of the current tradeoffs between using copper wire versus fiberoptic cable for rewiring buildings at NIST for data communications to the desk-top. We consider tradeoffs from two different viewpoints: (a) the cost of installing and maintaining the wiring, and (b) the computer imposed limitations on the use of the wiring. These findings substantiate our decision to use copper wiring.

Cooper, D.A.; A More Efficient Use of Delta-CRLs; Proceedings of the 2000 IEEE Symposium on Security and Privacy; 3/14/2000

Delta-CRLs were designed to provide a more efficient way to distribute certificate status information. However, as this paper will show, in some environments, if delta-CRLs are used as was originally intended, the benefits of using delta-CRLs will be minimal. This paper provides an analysis of delta-CRLs demonstrating the problems associated with issuing delta-CRLs in the "traditional" manner. A more efficient technique for issuing delta-CRLs is then presented.

Coriell, S.R., McFadden, G.B., Billia, B., Thi, H.N., Dabo, Y.; Electrical Pulsing during Directional Solidification: Analysis of Transients by Laplace Transform; Journal of Crystal Growth

The effect of the instantaneous application of an electric current during the directional solidification of a binary alloy is considered. The initial state consists of steady-state solidification at constant velocity. It is assumed that the electric current is sufficiently small that its application can be treated as a small perturbation. The problem is solved by Laplace transform techniques and the results are compared to numerical calculations for arbitrary currents. The Laplace transform allows for an understanding of the discontinuities resulting from the discontinuous electric pulse.

Coriell, S.R., McFadden, G.B., Murray, B.T.; Convective and Morphological Instabilities During Crystal Growth; Proceedings of the NASA Microgravity Materials Science Conference, Huntsville, Alabama, June 6-8, 2000; 6/6/2000

We have studied the effects of interface morphology on the dynamics of dendritic growth. The Ivantsov solution for an isothermal paraboloid of revolution growing into a pure, supercooled melt provides a relation between the bulk supercooling and a dimensionless product of the growth velocity and tip radius of a dendrite. In order to model dendritic growth in cubic materials, approximate solutions for paraboloids having perturbations with four-fold axial symmetry have been found. These solutions provide self-consistent corrections through second order in the shape parameter to the Peclet number-supercooling relation of the Ivantsov solution. The shape parameter is proportional to the amplitude of the four-fold correction to the dendrite shape, as measured from the Ivantsov paraboloid of revolution. We have calculated the shape parameter by comparing the dendrite tip shape to the portion of the equilibrium shape near the growth direction for anisotropic surface free energy with cubic symmetry. This comparison results in a shape parameter that is independent of the Peclet number. The calculated shape parameter is in good agreement with recent measurements for succinonitrile by LaCombe et al. This research has also provided analytic representations for the equilibrium shape of anisotropic crystals.

Courson, M.M., Marcais, G., Mink, A., Traverse, B.; An Automated Benchmarking Toolset; Springer Verlag Series Lecture Notes

The performance drive of parallel computing and platform upgrades or replacements are among the reasons frequent running of benchmark codes has become commonplace for application and platform evaluation and tuning. NIST is developing a prototype for an automated benchmarking toolset to reduce the manual effort in running and analyzing the results of such benchmarks. Our toolset consists of three main modules. A Data Collection and Storage module handles the collection of performance data and implements a central repository for such data. Another module provides an integrated mechanism to analyze and visualize the data stored in the repository. An Experiment Control Module assists the user in designing and executing experiments. To reduce the development effort, this toolset is built around existing tools and is designed to be easily extensible to support other tools.

Courson, M.M., Mink, A., Marcais, G., Traverse, B.; A Performance Measurement Environment for Cluster Computing; ISCA - DCS-2000 Conference

A number of factors promote frequent running of performance measurement codes for the evaluation and tuning of both applications and platforms. NIST is developing a prototype for an automated software measurement toolset to reduce the manual effort in running analyzing the results of such experiments. Our toolset consists of 3 main modules. A Data Collection and Storage module handles the collection of performance data and implements a central repository for such data. Another module provides an integrated mechanism to analyze and visualize the data stored in the repository. An Experiment Design and Control module assists the user in designing and executing experiments. To reduce the development effort, this toolset in built around existing tools and is designed to be easily extensible to support the addition.

Cugini, J.V.; Presenting Search Results: Design, Visualization, and Evaluation; Information Doors -- Where Information Search and Hypertext Link Workshop

A number of projects within the National Institute of Standards and Technology (NIST) have addressed the visual presentation and evaluation of search results. Within the design space for this problem, we distinguish between the logical structure imposed on the result set and the interface by which the structured results are presented to the user. This interface comprises the operations provided for the manipulation of the set as well as its visual presentation. Any design, no matter how intuitively appealing, should be evaluated and the full array of issues for HCI testing come into play. In particular, researchers must decide on a base case against which to measure, whether to use high-level and/or low-level metrics and which tasks are appropriate for the evaluation.

Cugini, J.V.; Web Usability Logging: Tools and Formats; Tools to Support Faster and Better Usability Engineering Workshop

Systematic usability studies of Web-based applications require software tools and data standards to support both the capture and representation of users' activities. Web logging tools can be based within the browser, the Website, or the underlying server. We compare the advantages and costs of these strategies. Logging tools produce log files, which must then be analyzed by the usability engineer. A common format for such log files would greatly facilitate information exchange and enable common software for presentation and analysis.

Dabo, Y., Thi, H.N., Coriell, S.R., McFadden, G.B., Billia, B.; On the Origin of Microsegregation in Peltier Interface Demarcation; Journal of Crystal Growth

Experimental results on solute microsegregation induced by Peltier Interface Demarcation (PID) technique during directional solidification of Bi - 1 wt % Sb alloys are presented. Using these data and numerical simulation, the theory of PID is revisited. It is shown that the limited growth kinetics has a dominant effect so that the commonly made assumption of local equilibrium at the solid-liquid interface (i.e., usual hypothesis of constant interface temperature during pulse marking for pure systems) should be abandoned, and the right dependence of interface temperature on solidification velocity included in the model. Consequently, the Peltier coefficients previously determined using Peltier pulsing have been more or less underestimated. Finally, two conditions to select systems capable of efficient marking by PID microsegregation are deduced and the effects of applied current in the first instants of electric pulse clarified.

Dabrowski, C.E., Huang, H., Messina, E., Horst, J.A.; Formalizing the NIST 4-D/RCS Reference Model Architecture Using an Architectural Description Language; NISTIR 6443; 12/31/1999

The 4-D/Real-Time Control System (RCS) Reference Model Architecture provides a well-defined strategy for development of software components for applications in robotics, automated manufacturing, and autonomous vehicles. An investigation has been conducted into the use of Architectural Description Languages (ADLs) as a means to provide a more formal, rigorous definition of the 4-D/RCS architecture and to specify software components for 4-D/RCS systems. In this report, we describe the results of an experiment into the use of an ADL to specify 4-D/RCS systems and assess the potential value of ADLs as specification and development tools for RCS domain experts. We conclude that ADLs not only can be used successfully to specify the 4-D/RCS Reference Model, but that they also serve as effective tools to enhance and extend this model. We also find that ADLs potentially can provide a formal basis for automatically checking the consistency of architecture specifications and verifying that designs conform to the reference architecture. This report discusses prospects for automated reuse of components specified with an ADL and makes recommendations on improving ADLs as effective tools for specifying, communicating, and validating 4-D/RCS system designs and software components. Finally, the report discusses potential influence of ADLs for commercial software development tools and provides future directions for research.

Davies, M.A., Burns, T.J.; Thermo-Mechanical Scillations in Material Flow During High-Speed Machining; Accepted by Philosophical Transactions of the Royal Society, London Series A

The chip formation process in machining involves a very high-strain-rate, high-temperature, nonlinear plastic flow of material. It has been demonstrated through careful observation that this plastic flow can exhibit at least five distinct dynamic flow patterns that are manifested in the following chip forms: (1) continuous; (2) continuous with built-up edge; (3) discontinuous; (4) shear localized; and (5) segmental (continuous with periodic variation in thickness). A unified approach to the problem would describe each type of chip as a stable dynamic equilibrium of the partial differential equations describing the elastic-plastic flow. This paper demonstrates this approach for the specific cases of continuous and shear localized chip formation. Both experiments and a simplified one-dimensional model of the flow show that, as cutting speed is increased, a transition takes place from continuous to shear-localized chip formation in the flow field of the material being cut. Initially, the process appears to be somewhat disordered. With further increases in cutting speed, the average spacing between shear bands increases monotonically, and the spacing becomes more regular and asymptotically approaches a limiting value that is determined by the cutting conditions and the properties of the work piece material. It is hoped that an analogous approach can be used to describe other transitions in the plastic flow field (chip form) as well.

Devaney, J.E.; The Role of Choice in Discovery; Lecture Notes in Computer Science

The discovery of functional forms can be divided into three tasks: 1 - choosing a search technique to find the set of best equations within the limitations of finite precision arithmetic and noise; 2 - choosing from among the best equations based on some criteria that encodes preference; and 3 - choosing a metric for the inductive support of the found equation. This paper looks at all three requirements and details the role of choice in each.

Devine, K., Hendrickson, B., Boman, E., St. John, M., Vaughan, C., Mitchell, W. F.; Zoltan: A Dynamic Load-Balancing Library for Parallel Applications Developer's Guide; Developer's Guide, Sandia Technical Report, SAND99-1376, 2000

The Zoltan Dynamic Load-Balancing Library provides critical capability to a number of parallel applications. Zoltan includes a suite of algorithms for dynamically computing partitions of problems over sets of processors; geometric, tree-based and graph-based algorithms are included. This document is a guide for developers of the Zoltan library.

Devine, K., Hendrickson, B.; Boman, E., St. John, M., Vaughan, C., Mitchell, W. F.; Zoltan: A Dynamic Load-Balancing Library for Parallel Applications User's Guide; User's Guide, Sandia Technical Report, SAND99-1377, 2000

The Zoltan Dynamic Load-Balancing Library provides critical capability to a number of parallel applications. Zoltan includes a suite of algorithms for dynamically computing partitions of problems over sets of processors; geometric, tree-based and graph-based algorithms are included. This document is a guide for users of the Zoltan library.

Dienstfrey, A., Huang, J., Hang, F.; Lattice Sums and the Two-Dimensional, Periodic Green's Function for the Helmholtz Equation; Accepted by the Proceedings of the Royal Society of London Mathematical, Physical and Engineering Sciences

Many algorithms that are currently used for the solution of the Helmholtz equation in periodic domains require the evaluation of the Green's function, ${\rm G}(\vec{x},\vec{x}_0)$. The fact that the natural representation of ${\rm G}$ via the method of images gives rise to a conditionally convergent series whose direct evaluation is prohibitive has inspired the search for more efficient procedures for evaluating this Green's function. Recently, the evaluation of ${\rm G}$ through the ``lattice sum'' representation has proven to be both accurate and fast. As a consequence, the computation of the requisite, also conditionally convergent, lattice sums has become an active area of research. In the following paper, we describe a new integral representation for these sums, and compare our results with other techniques for evaluating similar quantities.

Dimmick, D. L., Smith, Martin P.; Improving Weighting Schemes: Experiences Using Pivoted Document Length Normalization; Journal of Information Processing and Management

Today's state of the art information retrieval document ranking functions contain several tuning parameters. The proper settings for these parameters for a given collection may not be known and may be difficult to derive. This paper reports on efforts to derive a document ranking function with a single tuning parameter that can be set once and will work across many different collections. The performance target for this function is to meet or exceed the performance level of the top weighting functions as evaluated in the TREC conference. To achieve this goal experiments were carried out on several simple weighting functions using pivoted document length normalization to improve their performance. The results show that good performance can be achieved by using a simple pivoted normalization function that works well across the collections used in this study.

Dray, J.F.; Report on the NIST JavaT-AES Candidate Algorithm Analysis; Web page; 11/8/1999

NIST solicited candidate algorithms for the Advanced Encryption Standard in a Federal Register Announcement dated September 12, 1997. Fifteen candidates were submitted, and NIST has subsequently worked with a worldwide community of cryptanalysts and computer security researchers to analyze these candidates. This paper documents the test environment created by NIST for this purpose and the test results. Candidates were tested for memory usage, speed, and complexity. A final ranking of the candidates was obtained by normalizing the raw test results and calculating final composite scores that reflect performance across all test categories.