Expanded design procedures for learnable, usable interfaces (panel session)

ABSTRACT: This study examines how external auditors' accessibility to “tone at the top” knowledge impacts subsequent audit judgments. To examine this relationship, a decision aid is investigated that differentially facilitates the auditors' retrieval of “tone at the top” evidence from memory. Results of an experiment indicate that, holding the client's “tone at the top” constant, the structure of a control environment decision aid influences the auditors' mental representation of the “tone at the top.” Further, favorable “tone at the top” mental representations transfer to induce relatively favorable control environment and fraud risk assessments, and greater reliance on management's explanation for variances detected in analytical procedures. Mediation analyses identify the control environment assessment as a mediator between the influenced mental representation and the subsequent fraud risk and analytical procedure judgments. The results of the paper underscore the importance of how auditors develop their “tone at the top” mental representations, the influence of these mental representations on subsequent audit judgments, and the stage in the audit process where interventions can improve audit quality. Data Availability: Contact the author.

THINGS-data reflects three large-scale neuroimaging and behavioral datasets of object processing in humans, comprising densely sampled functional MRI and magnetoencephalographic recordings, as well as 4.70 million similarity judgments in response to thousands of photographic images for up to 1854 objects.

Predicate (PA) and numeric (NA) abstractions are the two principal techniques for software analysis. In this paper, we develop an approach to couple the two techniques tightly into a unified framework via a single abstract domain called NumPredDom. In particular, we develop and evaluate four data structures that implement NumPredDom but differ in their expressivity and internal representation and algorithms. All our data structures combine BDDs (for efficient propositional reasoning) with data structures for representing numerical constraints. Our technique is distinguished by its support for complex transfer functions that allow two-way interaction between predicate and numeric information during state transformation. We have implemented a general framework for reachability analysis of C programs on top of our four data structures. Our experiments on non-trivial examples show that our proposed combination of PA and NA is more powerful and more efficient than either technique alone.

Many applications that run on programmable data planes rely on approximate data structures, due to insufficient in-network memory. However, programming with approximate data structures is challenging because it requires (1) expertise in streaming algorithms to select the data structures that best match an application's requirements, (2) meticulous configuration to minimize approximation error while fitting within the hardware constraints, and (3) proficiency in the low-level P4 language. To address these issues, we propose NAP, a high-level network programming language. The core of NAP is the versatile approximate dictionary abstraction that captures a wide range of compact data structures, while allowing programmers to simply specify the kinds of error an application can tolerate. We demonstrate the language's expressiveness, conciseness, and efficiency through a variety of network applications, each compiling to P4 for the Intel Tofino in less than a second and featuring 25X--50X fewer lines of code compared to the P4 output. We evaluate an approximate stateful firewall written in NAP with real campus traffic, achieving performance consistent with the predicted accuracy.

Family pedigree data as required by genetics has proven to be difficult to analyze. This paper presents a data collection and structuring technique developed for the program PDGRE, which significantly improves on past computer application performances and produces a binary tree representation for the data. Data is collected in family units, mother, father, and offspring. The units are related by linked lists of those individuals who appear in more than one unit. The algorithm TRANS translates the links from relative pointers to actual memory locations. TRANS also locates those individuals who will be root nodes for the binary tree representations of the pedigree data. Algorithm BUILD creates the binary tree for any root passed to it. Each binary tree represents a subset of the original data set. The subsets prove more suited to analysis than the entire pedigree. The number of subsets is determined by the number of roots located by TRANS. The end result is data represented by a well known data structure at a significantly lower cost than previous applications.

“I didn't think I'd like group work, but I ended up in a good team and it was a great experience. But I feel if you end up in a bad team it can really get you down and will affect your mark.” This student response (University of Exeter, U.K.) is typical and telling. At least some of the time teams work, and the progression from “didn't think I'd like” to “great experience” happens often enough that those of us who use team work have come to expect it, at least part of the time. The ability to work well in teams is essential for our graduates. It cuts across all the questions — what, where, why and how we teach. However, not all of our students find working as a member of a team a natural and easy thing to do. Not only that, even though the student quoted in the preceding paragraph uses the phrase “good team” and the phrase “bad team” these and many other terms are not well defined. The Computer Science (CS) academic community regards group project work as an essential component of any degree; the discipline's professional societies world-wide emphasize project and group work as preparation for professional practice. Project work is recognized as having many educational and social benefits, in particular providing students with opportunities for active learning. Nevertheless, managing project work is problematic, because CS projects are: expensive , demanding considerable supervision as well as technical resources; complex , marrying design, human communication, human-computer interaction, and technology to satisfy objectives ranging from consolidation of technical skills through provoking insight into organizational practice, teamwork and professional issues, to inculcating academic discipline and presentation skills; continually demanding , set in the context of a rapidly changing technology which affects technical objectives and demands ever-evolving skills in both students and supervisors. In a young and changing discipline, some aspect of project work is questioned in almost every institution. The three panelists bring experience from four educational institutions in two countries. Louise Moses is Chair of the Department of Computer Science and Information Systems at Mount Union College. She has supervised team projects in extra-departmental service courses, classes provided for departmental majors, and inter-disciplinary courses. During the summer term of the previous three academic years she has been Honorary Visiting Researcher in the Department of Computer Science in the School of Engineering at The University of Exeter, Exeter, U.K. In that position she has worked as part of the management team for first year students in the first year project. James Caristi is a professor of mathematics and computer science at Valparaiso University. He was the 1990 recipient of the Sears Roebuck Award for Teaching Excellence and Campus Leadership, and is the 1999 recipient of the Distinguished Teaching Award from the Indiana Section of the Mathematical Association of America. He has been using teams in different ways in computer science classes at all levels for over 15 years. Sally Fincher is a Lecturer in the Computing Laboratory at the University of Kent at Canterbury in England. She has been project manager for the Effective Projectwork in Computer Science (EPCoS) project. EPCoS was a 10-partner, three-year funded project which worked to identify best practices in CS projectwork, transfer those practices between institutions and examine and analyze the process of transfer. We have wrestled with project design and how to make team assignments. And, even though providing good teamwork experiences is more of an art than a science — with no hard and fast rules — there are guidelines; it will be profitable to share our knowledge and our experiences. We shall consider six major areas concerning team and group work, and the kind of issues that are associated with them. Allocation How do we allocate students to groups? And then groups to supervisors? Supervision What sort of role should a supervisor take with respect to their group? Friend, mentor, project manager or technical guru? Does it make a difference? Assessment How do we assess the contribution of an individual when the deliverables are a team effort ? Should we even try to? Motivation What happens when students get into a “bad” team? How do we keep them motivated? Reflection Especially when introducing teamwork into the curriculum, reflection is an essential part of the learning cycle. How do we plan to make sure we include time and opportunity for this? Teamwork How do we encourage working together, when in some other academic circumstances this might be called “cheating”?

Algorithm animation attempts to explain an algorithm by visualizing interesting events of the execution of the implemented algorithm on some sample input. Algorithm explanation describes the algorithm on some adequate level of abstraction, states invariants, explains how important steps of the algorithm preserve the invariants, and abstracts from the input data up to the relevant properties. It uses a small focus onto the execution state. This paper is concerned with the explanation of algorithms on linked data structures. The thesis of the paper is that shape analysis of such algorithms produces abstract representations of such data structures, which focus on the active parts, i.e., the parts of the data structures, which the algorithm can access during it's next steps. The paper presents a concept of visually executing an algorithm on these abstract representations of data.

The chapter “Data Summaries and Representations: Definitions and Practical Use” examines data structures used to deal with complex networked data, using temporal networks as a concrete case. Complex networked data has become available in a variety of contexts, describing a variety of systems with growing abundance of details, such as, for instance, links between individuals in social networks, or the temporal evolution of these links. However, data needs to be summarized and represented in simple forms. This chapter describes several commonly used data summaries and levels of representation of temporal networks, as well as novel data representations that have been developed through the MULTIPLEX project. It focuses in particular on the case of temporal networks of contacts between individuals and shows in a series of concrete use cases how different representations can be used to characterize and compare data, or feed data-driven models of epidemic spreading processes.

Collections are commonly implemented as libraries by data structure experts, and are relied on heavily by application developers. The expert’s task is to implement a wide range of collections, and the application developer’s task is to pick an appropriate collection for each usage scenario. The design space for collections is huge, as data structures in practice implement not only their semantics, but also several performance-related concerns like memory layout and (im)mutability. This paper presents C♭, pronounced “C-flat”, a novel way to implement collections that lets experts implement the semantics of a collection data structure, in a way that is decoupled from its data representation. This simplifies collection implementation, and allows a collection’s performance to be tuned, for example, moving from a dense to a sparse representation, without changing its abstract specification. We describe C♭, both abstractly and in terms of a specific prototype implementation in Java. We use our prototype implementation to show that C♭ is expressive enough to implement common collections, that the code is straightforward, and that the performance of C♭ collections is close to Java’s standard collections for most operations, and much higher for some.

A level crossing predictor or alarm system with prediction horizon k is said to be optimal if it, at time t detects that an upcrossing will occur at time t + k, with a certain high probability and simultaneously gives a minimum number of false alarms. For a Gaussian stationary process, the optimal level crossing predictor can be explicitly specified in terms of the predicted value of the process itself and of its derivative. To the authors knowledge this simple optimal solution has not been used to any substantial degree. In this paper it is shown how a neural network can be trained to approximate an optimal alarm system arbitrarily well. As in other methods of parametrization, the choice of model structure, as well as an appropriate representation of data, are crucial for a good result. Comparative studies are presented for two Gaussian ARMA-processes, for which the optimal predictor can be derived theoretically. These studies confirm that a properly trained neural network can indeed approximate an optimal alarm system quite well – with due attention paid to the problems of model structure and representation of data. The technique is also tested on a strongly non-Gaussian Duffing process with satisfactory results.

Multidimensional trees

Logic simulation has become the bottleneck of today’s integrated circuit (IC) design projects. For instance, over 80 % of the IC design turn-around time of NVIDIA is spent on logic simulation even with NVIDIA’s proprietary supercomputing facility. It is thus essential to develop parallel simulation solutions to maintain the momentum of increasing IC integration capacity. Inspired by the supreme parallel computing power of modern GPUs, in this chapter we reported our recent work on using GPU to accelerate the time-consuming IC verification process by developing a massively parallel gate-level logical simulator. To the best of authors’ knowledge, this work is the first one to leverage the power of the modern GPUs to successfully unleash the massive parallelism of a conservative discrete event driven algorithm, CMB algorithm. Based on a novel data-parallel algorithmic mapping strategy, both the data structure and processing flow of the CMB protocol are re-designed to better exploit the potential of modern GPUs. A dynamic memory management mechanism is developed to efficiently utilize the relatively limited GPU memory resource. Experimental results prove that our GPU based simulator outperforms a CPU baseline event-driven simulator by a factor of 47.4X on average. This work demonstrates that the CMB algorithm can be efficiently and effectively deployed on GPUs without the performance overhead that had hindered its successful applications on previous parallel architectures.

A data structure is history independent if its internal representation reveals nothing about the history of operations beyond what can be determined from the current contents of the data structure. History independence is typically viewed as a security or privacy guarantee, with the intent being to minimize risks incurred by a security breach or audit. Despite widespread advances in history independence, there is an important data-structural primitive that previous work has been unable to replace with an equivalent history-independent alternative---dynamic partitioning. In dynamic partitioning, we are given a dynamic set S of ordered elements and a size-parameter B, and the objective is to maintain a partition of S into ordered groups, each of size Θ(B). Dynamic partitioning is important throughout computer science, with applications to B-tree rebalancing, write-optimized dictionaries, log-structured merge trees, other external-memory indexes, geometric and spatial data structures, cache-oblivious data structures, and order-maintenance data structures. The lack of a history-independent dynamic-partitioning primitive has meant that designers of history-independent data structures have had to resort to complex alternatives. In this paper, we achieve history-independent dynamic partitioning. Our algorithm runs asymptotically optimally against an oblivious adversary, processing each insert/delete with O(1) operations in expectation and O(B log N/loglog N) with high probability in set size N.

We present new geological, structural and microstructural data, together with K–Ar biotite and illite geochronological data, for the igneous–metamorphic rocks exposed in the Cuesta de Rahue Basement Inlier in the Southern Neuquén Precordillera, Argentina. We use these data to reconstruct the Late Paleozoic to Mesozoic tectonometamorphic and magmatic history of northwestern Patagonia. This block consists of a medium-grade metasedimentary sequence (the Cuesta de Rahue Metamorphic Complex), Late Carboniferous granitoids and a low-grade metavolcano-sedimentary unit (the Arroyo Coloco Metamorphic Complex). The Cuesta de Rahue Metamorphic Complex was deposited during the mid-Paleozoic and underwent Devonian low-pressure regional metamorphism, followed by the intrusion of granitoids at c. 300 Ma. By contrast, the Arroyo Coloco Metamorphic Complex records deformation and metamorphism under epizonal conditions (>300°C), constrained at c. 232–199 Ma by K–Ar and X-ray diffraction illite data. The Cuesta de Rahue Basement Inlier thus records a protracted orogenic evolution, recording Devonian metamorphism, Late Carboniferous–Permian Gondwanide tectonomagmatic processes, and Late Triassic deformation and metamorphism. This block was later affected by Mesozoic normal faulting and Miocene–Pliocene Andean deformation. The Andean deformation was intimately related to the reactivation of inherited basement fabrics, favouring a transpressional deformation regime. Supplementary material: Supplementary Material 1 – Analytical procedures is available at https://doi.org/10.6084/m9.figshare.c.6484467

Geographic information systems (GISs) are widely used for representation, management, and analysis of spatial data in many disciplines. In particular, geoscientists increasingly use these tools for data integration and management purposes in many environmental applications, ranging from water resources management to the study of global warming. Beyond these capabilities, geoscientists need to model and simulate three-dimensional (3D) dynamic fields and readily integrate those results with other relevant spatial information in order to have a better understanding of the environmental problems. However, GISs are very limited for the modeling and simulation of spatial fields, which are mostly 3D and dynamic. These limitations are mainly related to the existing GIS spatial data structures that are static and limited to 2D space. In order to overcome these limitations, we develop and implement a new kinetic 3D spatial data structure based on Delaunay tetrahedralization and a 3D Voronoi diagram to support a 3D dynamic field simulation within GISs. In this article, we describe in detail the different steps from discretization of a 3D continuous field to its numerical integration, based on an event-driven method. For validation of the proposed spatial data structure itself and its potential for the simulation of a dynamic field, two case studies are presented in the article. According to our observations, during the simulation process, the data structure is maintained and the 3D spatial information is managed adequately. Furthermore, the results obtained from both experiments are very satisfactory and are comparable with the results obtained from other existing methods for the simulation of the same dynamic field. To conclude, we discuss the current challenges related to the development of the 3D kinetic data structure itself and its adaptation to 3D dynamic field simulation and suggest some solutions for its improvement.