Glyphmaker: creating customized visualizations of complex data

Glyphmaker allows nonexpert users to customize their own graphical representations using a simple glyph editor and a point-and-click binding mechanism. In particular, users can create and then alter bindings to visual representations, bring in new data or glyphs with associated bindings, change ranges for bound data, and do these operations interactively. They can also focus on data down to any level of detail, including individual elements, and then isolate or highlight the focused region. These features empower users, letting them employ their specialized domain knowledge to create customized visual representations for further exploration and analysis. For ease of design and use, we built Glyphmaker on top of Iris Explorer, the Silicon Graphics Inc. (SGI) dataflow visualization system. The current version of Glyphmaker has been successfully tested on a materials system simulation. We are planning a series of tests and evaluations by scientists and engineers using real data     

Efficient and effective Querying by Image Content

In the QBIC (Query By Image Content) project we are studying methods to query large on-line image databases using the images' content as the basis of the queries. Examples of the content we use include color, texture, shape, position, and dominant edges of image objects and regions. Potential applications include medical (Give me other images that contain a tumor with a texture like this one), photo-journalism (Give me images that have blue at the top and red at the bottom), and many others in art, fashion, cataloging, retailing, and industry. We describe a set of novel features and similarity measures allowing query by image content, together with the QBIC system we implemented. We demonstrate the effectiveness of our system with normalized precision and recall experiments on test databases containing over 1000 images and 1000 objects populated from commercially available photo clip art images, and of images of airplane silhouettes. We also present new methods for efficient processing of QBIC queries that consist of filtering and indexing steps. We specifically address two problems: (a) non Euclidean distance measures; and (b) the high dimensionality of feature vectors. For the first problem, we introduce a new theorem that makes efficient filtering possible by bounding the non-Euclidean, full cross-term quadratic distance expression with a simple Euclidean distance. For the second, we illustrate how orthogonal transforms, such as Karhunen Loeve, can help reduce the dimensionality of the search space. Our methods are general and allow some false hits but no false dismissals. The resulting QBIC system offers effective retrieval using image content, and for large image databases significant speedup over straightforward indexing alternatives. The system is implemented in X/Motif and C running on an RS/6000.On sabbatical from Univ. of Maryland, College Park. His work was partially supported by SRC, by the National Science Foundation under the grant IRI-8958546 (PYI).     

Engineering insightful visualizations

A theoretical visualization model that is suitable for a guideline based engineering approach as well as generically and widely applicable to visualization and its subfields is developed in this work. It is based on investigating question–answer pairs and emphasizes understanding and knowledge acquisition achieved via insight and learning but which is impeded by confusion brought on by in-appropriateness, incoherence, anacolutha and non-sequiturs. These terms are technically defined within the model. A visualization metric is developed that relates insight, learning and confusion with characteristics of how much and how fast understanding and knowledge are acquired. The model entails two connected processes: a visualization process based on visualization media componentization followed by a human process consisting of perception, interpretation, understanding and knowledge acquisition. Several case studies drawn from the various subfields of visualization show the potential of the proposed model.     

Proof by analogy in mural

An important advantage of using a formal method of developing software is that one can prove that development steps are correct with respect to their specification. Conducting proofs by hand, however, can be time consuming to the extent that designers have to judge whether a proof of a particular obligation is worth conducting. Even if hand proofs are worth conducting, how do we know that they are correct? One approach to overcoming this problem is to use an automatic theorem proving system to develop and check our proofs. However, in order to enable present day theorem provers to check proofs, one has to conduct them in much more detail than hand proofs. Carrying out more detailed proofs is of course more time consuming. This paper describes the use of proof by analogy in an attempt to reduce the time spent on proofs. We develop and implement a proof follower based on analogy and present an example to illustrate its characteristics. The example shows that even when the follower fails to complete a proof, it can provide a hint that enables the user to complete a proof.     

Constrained recommendations for query visualizations

The improvement of data storage and data acquisition techniques has led to huge accumulated data volumes in a variety of applications. International research enterprises such as the Human Genome and the Digital Sky Survey Projects are generating massive volumes of scientific data. A major challenge with these datasets is to glean insights from them to discover patterns or to originate relationships. The analysis of these massive, typically messy, and inconsistent volumes of data is indeed crucial and challenging in many application domains. Hence, the research community has introduced a number of visualizations tools to guide and help analysts in exploring the data space to extract potentially useful information. However, when working with high-dimensional datasets, identifying visualizations that show interesting variations and trends in data is not trivial: the analyst must manually specify a large number of visualizations, explore relationships among various attributes, and examine different subsets of data before discovering visualizations that are interesting or insightful. Though, exploring all possible visualizations involves complex challenges. It is a costly and time-consuming process especially when the dimensionality is high. Furthermore, the rapid growth of databases becomes multifaceted in their channels and dimensionality; thus, the transition from static analysis to real-time analytics represents a fundamental paradigm shift in the field of Big Data. Motivated by the above challenges, we propose an efficient framework called real-time scoring engine (RtSEngine) that assists analysts to limit the exploration of visualizations for a specified number of visualizations and/or certain execution time quote to recommend a set of visualizations that meet analysts’ budgets. To achieve that, RtSEngine incorporates our proposed approaches to prioritize and score attributes that form all possible visualizations in a dataset based on their statistical properties such as selectivity, data distribution, and number of distinct values. Then, RtSEngine recommends the visualizations created from the top-scored attributes. Moreover, we present visualizations cost-aware techniques that estimate the retrieval and computation costs of each visualization so that analysts may discard high-cost visualizations. We show and evaluate the effectiveness and efficiency of our proposed approaches, and asses the quality of visualizations and the overhead obtained by applying our techniques on both synthetic and real datasets.     

Grid-distributed visualizations using connectionless protocols

The NERSC and Lawrence Berkeley National Laboratory visualization group has developed the Visapult tool to attack grand challenge problems. Visapult is a distributed, parallel, volume rendering application that leverages parallel computation and high-performance networking resources that are on the same scale as the supercomputers generating the data. We've improved Visapult's effectiveness using aggressive network tuning and network protocol modifications. In particular, we used a new connectionless user datagram protocol (UDP) to improve network efficiency from a 25 to 88 percent line rate increase for multigigabit networks. This connectionless protocol also dramatically reduces the latency of network event delivery, improving the responsiveness of wide area distributed interactive graphics applications as compared to transmission control protocol (TCP) streams. We believe that this UDP protocol, as well as transport encodings and algorithms that can tolerate loss gracefully, will become a fundamental component of future grid visualization architectures.     

The problem of reasoning by analogy

This article is the twenty-seventh of a series of articles discussing various open research problems in automated reasoning. The problem proposed for research asks one to find criteria that an automated reasoning program can apply to determine whether to attack a given question with reasoning by analogy. The imprecise term reasoning by analogy refers to a type of reasoning in which the type of proof being sought is sharply influenced by the style of proof that was successfully used to prove related theorems.This work was supported by the Office of Scientific Computing, U.S. Department of Energy, under Contract W-31-109-Eng-38.     

Querying now-relative data

Now-relative temporal data play an important role in most temporal applications, and their management has been proved to impact in a crucial way the efficiency of temporal databases. Though several temporal relational approaches have been developed to deal with now-relative data, none of them has provided a whole temporal algebra to query them. In this paper we overcome such a limitation, by proposing a general algebra which is parametrically adapted to cope with the relational approaches to now-relative data in the literature, i.e., MIN, MAX, NULL and POINT approaches. Besides being general enough to provide a query language for several approaches in the literature, our algebra has been designed in such a way to satisfy several theoretical and practical desiderata: closure with respect to representation languages, correctness with respect to the “consensus” BCDM semantics, reducibility to the standard non-temporal algebra (which involves interoperability with non-temporal relational databases), implementability and efficiency. Indeed, the experimental evaluation we have drawn on our implementation has shown that only a slight overhead is added by our treatment of now-relative data (with respect to an approach in which such data are not present).     

Querying Linguistic Trees

Large databases of linguistic annotations are used for testing linguistic hypotheses and for training language processing models. These linguistic annotations are often syntactic or prosodic in nature, and have a hierarchical structure. Query languages are used to select particular structures of interest, or to project out large slices of a corpus for external analysis. Existing languages suffer from a variety of problems in the areas of expressiveness, efficiency, and naturalness for linguistic query. We describe the domain of linguistic trees and discuss the expressive requirements for a query language. Then we present a language that can express a wide range of queries over these trees, and show that the language is first-order complete over trees.     

Querying XML streams

Efficient querying of XML streams will be one of the fundamental features of next-generation information systems. In this paper we propose the TurboXPath path processor, which accepts a language equivalent to a subset of the for-let-where constructs of XQuery over a single document. TurboXPath can be extended to provide full XQuery support or used to augment federated database engines for efficient handling of queries over XML data streams produced by external sources. Internally, TurboXPath uses a tree-shaped path expression with multiple outputs to drive the execution. The result of a query execution is a sequence of tuples of XML fragments matching the output nodes. Based on a streamed execution model, TurboXPath scales up to large documents and has limited memory consumption for increased concurrency. Experimental evaluation of a prototype demonstrates performance gains compared to other state-of-the-art path processors.Received: 30 January 2003, Accepted: 4 February 2004, Published online: 8 April 2004Edited by: R. Baeza-Yates.     

半结构化数据查询的处理和优化

半结构化数据的特点是数据的结构不规则或不完整,其模型都基于带根有向图,因此,查询处理过程本质上是对图的搜索过程.另外,通配路径使查询处理更加复杂化.文章详细介绍了异构数据源集成系统Versatile中采取的半结构数据OIM(model for object integration)对象的查询和优化策略,包括查询计划的生成、路径扩展和路径索引、层次索引和基于数据源知识这3种查询优化方法.文章介绍的方法同样适用于其他的半结构化数据模型.     

VisLink: revealing relationships amongst visualizations

We present VisLink, a method by which visualizations and the relationships between them can be interactively explored. VisLink readily generalizes to support multiple visualizations, empowers inter-representational queries, and enables the reuse of the spatial variables, thus supporting efficient information encoding and providing for powerful visualization bridging. Our approach uses multiple 2D layouts, drawing each one in its own plane. These planes can then be placed and re-positioned in 3D space: side by side, in parallel, or in chosen placements that provide favoured views. Relationships, connections, and patterns between visualizations can be revealed and explored using a variety of interaction techniques including spreading activation and search filters.     

A probabilistic approach to pronunciation by analogy

The relationship between written and spoken words is convoluted in languages with a deep orthography such as English and therefore it is difficult to devise explicit rules for generating the pronunciations for unseen words. Pronunciation by analogy (PbA) is a data-driven method of constructing pronunciations for novel words from concatenated segments of known words and their pronunciations. PbA performs relatively well with English and outperforms several other proposed methods. However, the method inherently generates several candidate pronunciations and its performance depends critically on a good scoring function to choose the best one of them.Previous PbA algorithms have used several different scoring heuristics such as the product of the frequencies of the component pronunciations of the segments, or the number of different segmentations that yield the same pronunciation, and different combinations of these methods, to evaluate the candidate pronunciations. In this article, we instead propose to use a probabilistically justified scoring rule. We show that this principled approach alone yields better accuracy than any previously published PbA algorithm. Furthermore, combined with certain ad hoc modifications motivated by earlier algorithms, the performance can in some cases be further increased.     

Querying large models efficiently

ContextThe paradigm of Model-Driven Engineering (MDE) has emerged as a new area of software engineering that uses models to improve the productivity and reusability of software in order to achieve industrial standards. As models grow in size and complexity, the need of model persistence and model querying solutions arises to efficiently store large models and obtain information from them in an efficient, usable and safe way. Morsa is a model repository that uses a No-SQL database backend; it has been recently presented [1] and achieves scalable access to models and transparent integration with tools.ObjectiveOur goal was to develop a query language for Morsa, as the existing model querying approaches cannot take advantage of the design of the our repository.MethodThe method followed in this paper comprises the following steps: (i) analyze the problem of model querying and identify a set of dimensions that can be used to characterize querying approaches; (ii) study and evaluate a representative set of model querying approaches and (iii) use the experience gained to design, develop and evaluate a dedicated model querying approach for Morsa that performs better than the studied ones (plain EMF, EMF Query, MDT OCL, IncQuery and CDO OCL). A test case has been defined to evaluate and compare the different approaches.ResultsThe contributions of this work are: first, an efficient, usable querying approach called Morsa Query Language (MorsaQL) that extends Morsa with querying capabilities, as the existing querying approaches cannot take advantage of our repository, and second, a comparative study of the current model persistence and querying approaches.ConclusionThe experience of analyzing and evaluating the different querying approaches has been very useful, as it has helped us developing our own one, which has been proven to be the best choice for Morsa. Moreover, the results of this paper can guide the MDE developers on which querying approach to use, depending on their needs.     

本体的查询与推理研究

现今,计算机网络被广泛应用于生活的方方面面,而从海量的信息中搜寻出人们所需要的还存在诸多问题,于是产生了本体的概念。而本体的查询和推理是基于本体的应用中重要的组成部分,研究的目的是为了使知识得以充分表达并且对信息的查询更加精确、完备。首先介绍了本体的概念并建立本体模型,然后用本体杳询语言SPARQL对已有模型进行查询并用SWRL对模型进行语义规则的扩充;最后介绍了Jena,并对本体模型进行推理,由此获得了更多知识。结论就足,在利用SPARQL和Jena进行查询与推理的过程中,推理将提高查询能力,而规则是提高推理能力的关键。     

Querying Semistructured Heterogeneous Information

Semistructured data has no absolute schema fixed in advance and its structure may be irregular or incomplete. Such data commonly arises in sources that do not impose a rigid structure (such as the World-Wide Web) and when data is combined from several heterogeneous sources. Data models and query languages designed for well structured data are inappropriate in such environments. Starting with a lightweight object model adopted for the TSIMMIS project at Stanford, in this paper we describe a query language and object repository designed specifically for semistructured data. Our language provides meaningful query results in cases where conventional models and languages do not: when some data is absent, when data does not have regular structure, when similar concepts are represented using different types, when heterogeneous sets are present, and when object structure is not fully known. This paper motivates the key concepts behind our approach, describes the language through a series of examples (a complete semantics is available in an accompanying technical report [23]), and describes the basic architecture and query processing strategy of the lightweight object repository we have developed.     

Artistic collaboration in designing VR visualizations

This article describes one of the recent major collaborative efforts, a class on designing VR scientific visualizations that was co-taught with professors and students from Brown University's computer science department and the Rhode Island School of Design's (RISD's) illustration department. We discuss here the experiences that led us to this conclusion; along with some of the tools we have developed to facilitate working with artists. Many of the experiences and conclusions relayed here are the results of this class. We then discuss three important themes that we derived from our experiences, which are all motivated by a desire to better facilitate artistic collaborations.     

Model-driven visualizations of constraint-based local search

Visualization is often invaluable to understand the behavior of optimization algorithms, identify their bottlenecks or pathological behaviors, and suggest remedial techniques. Yet developing visualizations is often a tedious activity requiring significant time and expertise. This paper presents a framework for the visualization of constraint-based local search (CBLS) algorithms. Given a high-level model and a declarative visualization specification, the CBLS visualizer systematically produces animations to visualize constraints and objectives, violations, and conflicts, as well as the temporal behavior of these measures. The visualization specification is declarative and typically composed of a triple (what,where,how) indicating what to display, where, and with which graphical objects. The visualizer architecture is compositional and extensible. It provides building blocks which can be assembled freely by the user and focuses almost exclusively on static aspects, the dynamic aspects being automated by the use of invariants. The paper highlights various functionalities of the visualizer and describes a blueprint for its implementation.