Grounded Symbolic Communication between Heterogeneous Cooperating Robots

In this paper, we describe the implementation of a heterogeneous cooperative multi-robot system that was designed with a goal of engineering a grounded symbolic representation in a bottom-up fashion. The system comprises two autonomous mobile robots that perform cooperative cleaning. Experiments demonstrate successful purposive navigation, map building and the symbolic communication of locations in a behavior-based system. We also examine the perceived shortcomings of the system in detail and attempt to understand them in terms of contemporary knowledge of human representation and symbolic communication. From this understanding, we propose the Adaptive Symbol Grounding Hypothesis as a conception for how symbolic systems can be envisioned.     

Multiple Comparisons in Induction Algorithms

A single mechanism is responsible for three pathologies of induction algorithms: attribute selection errors, overfitting, and oversearching. In each pathology, induction algorithms compare multiple items based on scores from an evaluation function and select the item with the maximum score. We call this a multiple comparison procedure (MCP). We analyze the statistical properties of MCPs and show how failure to adjust for these properties leads to the pathologies. We also discuss approaches that can control pathological behavior, including Bonferroni adjustment, randomization testing, and cross-validation.     

Strategies for Expressing Concise, Helpful Answers

An intelligent help system needs to take into account the user's knowledgewhen formulating answers. This allows the system to provide more conciseanswers, because it can avoid telling users things that they already know.Since these concise answers concentrate exclusively on pertinent newinformation, they are also easier to understand. Information about theuser's knowledge also allows the system to take advantage of the user'sprior knowledge in formulating explanations. The system can provide betteranswers by referring to the user's prior knowledge in the explanation(e.g., through use of similes). This process of refining answers is calledanswer expression.The process of answer expression has been implemented in the UCExpresscomponent of UC (UNIX Consultant), a natural language system that helps theuser solve problems in using the UNIX operating system. UCExpress separatesanswer expression into two phases: pruning and formatting.In the pruning phase, subconcepts of the answer are pruned by being markedas already known by the user (and hence do not need to be generated), ormarked as candidates for generating anaphora or ellipsis (since they arepart of the conversational context). In the formatting phase, UCExpressuses information about the user's prior domain knowledge to select amongspecialized expository formats,such as similes and examples, for expressing information to the user. Theseformats allow UCExpress to present different types of information to theuser in a clear, concise manner. The result of UCExpress' answer expressionprocess is an internal form that a tactical level generator can easily useto produce good English.     

Decision-Centric Information Monitoring

We are interested in information management for decision support applications, especially those that monitor distributed, heterogeneous databases to assess time-critical decisions. Users of such applications can easily be overwhelmed with data that may change rapidly, may conflict, and may be redundant. Developers are faced with a dilemma: either filter out most information and risk excluding critical items, or gather possibly irrelevant or redundant information, and overwhelm the decision maker. This paper describes a solution to this dilemma called decision-centric information monitoring (DCIM). First, we observe that decision support systems should monitor only information that can potentially change some decision. We present an architecture for DCIM that meets the requirements implied by this observation. We describe techniques for identifying the highest value information to monitor and techniques for monitoring that information despite autonomy, distribution, and heterogeneity of data sources. Finally, we present lessons learned from building LOOKOUT, which is to our knowledge the first implementation of a top-to-bottom system performing decision-centric information monitoring.     

Compact modeling of amorphous‐silicon thin‐film transistors with BSIM3

Abstract— A novel approach of modeling a‐Si:H TFTs with the industry‐standard BSIM3 compact model is presented. The described approach defines the a‐Si:H TFT drain current and terminal charges as explicit functions of terminal voltages using a minimum set of BSIM3 parameters. The set of BSIM3 parameters is chosen based on the electrical and physical characteristics of the a‐Si:H TFT and their values extracted from measured data. By using the selected BSIM3 model parameters, the a‐Si:H TFT is simulated inside SPICE to fit the simulated I‐V and C‐V curves with the measured results. Finally, the extracted BSIM3 model is validated by simulating the kickback voltage effect in an AMLCD pixel array.     

Report on the 2nd International Summer School on Network and Service Management (ISSNSM’08)

This report summarizes the 2nd International Summer School on Network and Service Management (ISSNSM’08), which was held at the Communication Systems Group (CSG) of the Department of Informatics (IFI), University of Zurich, Switzerland, on 2–6 June 2008. Supported by the European FP6 Network of Excellence for the Management of Internet Technologies and Complex Services (EMANICS), the ISSNSM presented within 5 days eight different topics, covering the areas of (1) security, (2) virtualization and simulations, and (3) network monitoring and management. All of these run for a full or half day, including a short theoretical introduction and larger practical lab course components, respectively.

Using visual design experts in critique-based evaluation of 2D vector visualization methods

We describe an experiment in which art and illustration experts evaluated six 2D vector visualization methods. We found that these expert critiques mirrored previously recorded experimental results; these findings support that using artists, visual designers and illustrators to critique scientific visualizations can be faster and more productive than quantitative user studies. Our participants successfully evaluated how well the given methods would let users complete a given set of tasks. Our results show a statistically significant correlation with a previous objective study: designers' subjective predictions of user performance by these methods match users measured performance. The experts improved the evaluation by providing insights into the reasons for the effectiveness of each visualization method and suggesting specific improvements.     

Volumetric Curved Planar Reformation for virtual endoscopy

Curved Planar Reformation (CPR) has proved to be a practical and widely used tool for the visualization of curved tubular structures within the human body. It has been useful in medical procedures involving the examination of blood vessels and the spine. However, it is more difficult to use it for large, tubular, structures such as the trachea and the colon because abnormalities may be smaller relative to the size of the structure and may not have such distinct density and shape characteristics. Our new approach improves on this situation by using volume rendering for hollow regions and standard CPR for the surrounding tissue. This effectively combines gray scale contextual information with detailed color information from the area of interest. The approach is successfully used with each of the standard CPR types and the resulting images are promising as an alternative to virtual endoscopy. Because the CPR and the volume rendering are tightly coupled, the projection method used has a significant effect on properties of the volume renderer, such as distortion and isometry. We describe and compare the different CPR projection methods and how they affect the volume rendering process. A version of the algorithm is also presented which makes use of importance driven techniques; this ensures the users attention is always focused on the area of interest and also improves the speed of the algorithm.     

A comparison of the perceptual benefits of linear perspective and physically-based illumination for display of dense 3D streamtubes

Large datasets typically contain coarse features comprised of finer sub-features. Even if the shapes of the small structures are evident in a 3D display, the aggregate shapes they suggest may not be easily inferred. From previous studies in shape perception, the evidence has not been clear whether physically-based illumination confers any advantage over local illumination for understanding scenes that arise in visualization of large data sets that contain features at two distinct scales. In this paper we show that physically-based illumination can improve the perception for some static scenes of complex 3D geometry from flow fields. We perform human-subjects experiments to quantify the effect of physically-based illumination on participant performance for two tasks: selecting the closer of two streamtubes from a field of tubes, and identifying the shape of the domain of a flow field over different densities of tubes. We find that physically-based illumination influences participant performance as strongly as perspective projection, suggesting that physically-based illumination is indeed a strong cue to the layout of complex scenes. We also find that increasing the density of tubes for the shape identification task improved participant performance under physically-based illumination but not under the traditional hardware-accelerated illumination model.     

VisComplete: automating suggestions for visualization pipelines

Building visualization and analysis pipelines is a large hurdle in the adoption of visualization and workflow systems by domain scientists. In this paper, we propose techniques to help users construct pipelines by consensus--automatically suggesting completions based on a database of previously created pipelines. In particular, we compute correspondences between existing pipeline subgraphs from the database, and use these to predict sets of likely pipeline additions to a given partial pipeline. By presenting these predictions in a carefully designed interface, users can create visualizations and other data products more efficiently because they can augment their normal work patterns with the suggested completions. We present an implementation of our technique in a publicly-available, open-source scientific workflow system and demonstrate efficiency gains in real-world situations.