SBV:基于SVG的生物信息可视化软件

生物信息可视化是从生物大数据中挖掘有效信息的重要手段。针对生物信息的海量性、可视化效果的精确性、各种可视化需求的多样性等挑战,设计并实现了一款基于SVG矢量图的生物信息可视化软件SBV (SVG for Bioinformatics Visualization)。SBV充分利用了SVG的可伸缩性、DOM和CSS表现形式的可定制性,实现了10余种常用的生物信息用图,可支持现有的大部分生物信息可视化,是一款易于操作的综合型生物信息画图软件。目前该软件已经在Github上开源,为后续开发 更多功能奠定了较好的基础。     

3S技术在生物多样性研究中的应用

随着生物多样性问题的日益突出,对生物多样性的研究与保护提出了更高的要求。但由于传 统调查方法的局限性,迫切需要新的技术支持。将3S技术应用于生物多样性研究,一方面可以弥补 传统调查手段的不足,另一方面为生物多样性的调查研究提供了新的途径与技术支持。在总结3S 技术在生物多样性研究中的应用现状的基础上,对其应用发展前景进行了展望,指出了3S技术在 生物多样性研究中的系统化应用正朝着数据表达的可视化、综合研究和网络化、构件化以及智能化 的方向发展。     

分子结构与基因序列数据综合可视化方法研究

随着系统生物学的兴起和迅速发展,为了探索不同生物层次信息之间的关联关系,对不同层次生物科学计算数据综合可视化的需求日益迫切。不同层次的生物数据的组织管理是实现综合可视化的基础和关键技术,因此文中面向生物信息可视化领域的需求,基于综合可视化集成框架,提出了一种分子结构与基因序列数据的元数据组织模型与关联数据自动生成方法。该方法通过定义包括分子结构、基因序列以及分子结构/基因序列关联信息三类数据的元数据模型,建立了该两个层次和领域数据的关联关系,确定了关联数据描述格式;利用先进的XML技术实现了分子领域和基因序列领域元数据的自动提取和转换。在此基础之上开发了一个分子结构数据与基因序列数据综合可视化原型系统,取得了良好的试用效果。     

基于径向坐标可视化的高维数据分析方法

针对模式分类算法不直观的问题,提出一种基于径向坐标可视化分析高维数据的方法。由最大似然原理估计高维数据的本征维数,用较少的变量结合径向坐标可视化方法对高维数据进行可视化降维分析。在径向坐标中揭示高维数据集中类别和特征间的关系,寻找基于不同特征排列顺序的最优映射,并结合多种机器学习方法对数据集进行分类。应用于UCI数据库中的6个数据集的结果表明,该方法具有较好的可视化和分类效果。     

基于空间图层的交互式空间数据可视化研究

空间数据是具有可视化特性的"图形或图像"数据,迄今为止,很少有人关注空间数据的图层显示方式和空间数据的图形表达方式的结合。为此,本文首先实现了三种常用空间数据可视化工具,并选用MO软件包做为图层处理工具,在此基础上设计并实现了一系列可视化交互功能,以耦合图层显示和数据显示两种表达方式,最终达到了交互式数据可视化方便直观的效果。     

基于多维数据雷达图表示的图形分类器研究

分析介绍了用雷达图表示多维数据以及雷达图的图形特征选取和融合的基本方法,提出了一种基于多维数据雷达图表达的可视化图形分类新方法,该方法用雷达图表示多维数据,不同类别的多维数据对应不同的雷达图形,形成以雷达图形特征为表达主要特征的分类方法。并通过模糊推理方来来自动识别雷达图形,完成自动分类。实验表明,此方法简单、直观,实现了分类过程可视化、分类结果可视化,并且有良好的分类效果。     

基于SSM框架的社区环境数据可视化系统

为了更高效地对社区的环境数据进行管理和直观的展示,提高社区管理系统的使用效率,该文以PM2.5、O3、SO2等社区的环境数据为对象,基于Spring,SpringMVC和Mybatis框架设计了社区环境数据可视化管理系统.前端以Bootstrap为框架,ECharts作为可视化组件实现系统的主要功能.结果表明,该数据可视化系统具有性能良好、数据展示直观、代码复用率高的特性.     

数据挖掘可视化技术应用与研究

本文讨论了数据挖掘和可视化的关键技术,提出了运用在油田数据库中的一种可视化模型的设计方法。取出油田数据库的一个数据截面进行可视化,可以直观清晰的看到数据库中的频繁与例外异常模式,对提高决策的效率具有重大现实意义。     

P2P流量的可视化分析

详细介绍了在VC.NET平台下基于WinPcap和Mapx开发的P2P流量检测系统,将捕获的数据进行分组,分析出其流量特征,并以可视化的方式在地图背景上显示IP地址,以有向边的方式显示数据传输的路径。在现有P2P流量检测方法的基础上与地理信息系统结合,提出了一种具有代价小、效率高、直观特点的可视化分析方法,并给出实验数据对比说明。     

中文专利数据可视化分析系统研究与设计

现有的专利分析系统在可视化方面存在诸多不足，针对中文专利的可视化分析工具更是少之又少。结合文本挖掘技术，设计并实现“中文专利数据可视化分析系统”,具有专利态势分析、专利聚类分析、专利引证分析三种分析功能，多角度分析了中文专利数据，并将数据分析结果通过可视化模块直观地呈现出来，为专利信息使用者提供更好服务。     

ConceptGraph: A Formal Model for Interpretation and Reasoning During Visual Analysis

In order to discuss the kinds of reasoning a visualization supports and the conclusions that can be drawn within the analysis context, a theoretical framework is needed that enables a formal treatment of the reasoning process. Such a model needs to encompass three stages of the visualization pipeline: encoding, decoding and interpretation. The encoding details how data are transformed into a visualization and what can be seen in the visualization. The decoding explains how humans construct graphical contexts inside the depicted visualization and how they interpret them assigning meaning to displayed structures according to a formal reasoning strategy. In the presented model, we adapt and combine theories for the different steps into a unified formal framework such that the analysis process is modelled as an assignment of meaning to displayed structures according to a formal reasoning strategy. Additionally, we propose the ConceptGraph, a combined graph-based representation of the finite-state transducers resulting from the three stages, that can be used to formalize and understand the reasoning process. We apply the new model to several visualization types and investigate reasoning strategies for various tasks.     

An approach to the perceptual optimization of complex visualizations

This paper proposes a new experimental framework within which evidence regarding the perceptual characteristics of a visualization method can be collected, and describes how this evidence can be explored to discover principles and insights to guide the design of perceptually near-optimal visualizations. We make the case that each of the current approaches for evaluating visualizations is limited in what it can tell us about optimal tuning and visual design. We go on to argue that our new approach is better suited to optimizing the kinds of complex visual displays that are commonly created in visualization. Our method uses human-in-the-loop experiments to selectively search through the parameter space of a visualization method, generating large databases of rated visualization solutions. Data mining is then used to extract results from the database, ranging from highly specific exemplar visualizations for a particular data set, to more broadly applicable guidelines for visualization design. We illustrate our approach using a recent study of optimal texturing for layered surfaces viewed in stereo and in motion. We show that a genetic algorithm is a valuable way of guiding the human-in-the-loop search through visualization parameter space. We also demonstrate several useful data mining methods including clustering, principal component analysis, neural networks, and statistical comparisons of functions of parameters.     

VIS+AI: integrating visualization with artificial intelligence for efficient data analysis

Visualization and artificial intelligence (AI) are well-applied approaches to data analysis. On one hand, visualization can facilitate humans in data understanding through intuitive visual representation and interactive exploration. On the other hand, AI is able to learn from data and implement bulky tasks for humans. In complex data analysis scenarios, like epidemic traceability and city planning, humans need to understand large-scale data and make decisions, which requires complementing the strengths of both visualization and AI. Existing studies have introduced AI-assisted visualization as AI4VIS and visualization-assisted AI as VIS4AI. However, how can AI and visualization complement each other and be integrated into data analysis processes are still missing. In this paper, we define three integration levels of visualization and AI. The highest integration level is described as the framework of VIS+AI, which allows AI to learn human intelligence from interactions and communicate with humans through visual interfaces. We also summarize future directions of VIS+AI to inspire related studies.     

基于SOM聚类的垃圾处理方式可视化分析方法

针对全国各省份垃圾处理方式的数据,提出一种混合可视分析方法。为了从多角度分析数据,混合U矩阵、平行坐标以及Small-Multiple 3种可视化技术,设计并实现了3种可视化视图的交互联动。首先,对数据进行聚类处理,将各省份近年的垃圾处理方式划分类别,采用SOM神经网络聚类算法实现聚类。然后,针对SOM聚类结果,采用U矩阵的方式进行可视化,并采用平行坐标描述每个聚类结果的各个属性。为了分析数据的地理属性及时序属性,采用Small-Multiple可视化技术。最后,实现多视图联动、刷新技术等交互方式,帮助用户自行探索数据,实现多视图的交互展示与分析。实验表明,这种混合可视方式可达到较好的多属性交互可视化效果,能够帮助用户了解并分析我国垃圾处理方式的分布及趋势。     

基于本体可视化的关联数据集表达

近年来不同学科领域的语义数据井喷式增长,关联数据集关联关系的发布形式直接决定了这些数据集能否被快速、准确地理解和使用。目前,数据集关联关系的发布均采用文本,而本论文提出一种基于本体可视化的方法来表达关联数据集,使数据集更容易被理解和使用。设计关联数据集发布方案,基于 OWL2VOWL 项目实现将关联关系转化为 WebVOWL 规定格式的 JSON 元素,并将其结合 WebVOWL 嵌入到发布模型中,完成关联数据集关联关系的发布。依据以微生物为核心的关联数据集的关系可视化表示,完成在数据集上的查询,验证方案带来的便利。通过设计的本体可视化模型来完成关联数据集的表达,丰富了关联数据集的发布形式,并使得关联数据集更容易被专家甚至不懂本体的使用者理解和使用。     

Research on environmental landscape design based on virtual reality technology and deep learning

Virtual Reality (VR) provides immersive visualization and intuitive interaction. The VR is used to enable any biomedical profession to develop a deep learning (DL) model for image classification. The Deep Neural Network (DNN) models can be used as a powerful tool for data analysis, but they are also challenging to understand and develop. To make deep learning more convenient and fast operation, it have established a landscape of DNN development environment based on virtual reality. In this environment, users can move concrete objects only to build neural networks with their own hands. It automatically transforms these configurations into a trainable model and reports on the real-time test dataset. In addition to realizing the insights users are developing into DNN models, it has also visually enriched the virtual environmental landscape objects with the parts of the model. In this way bridge the gap between professionals in different disciplines, providing a new perspective on the model analysis and data interaction. This system further demonstrates that learning and visualization technologies developed in Shenzhen can benefit from integrating virtual reality.     

Worm plots

Scientists sometimes spend hours conducting experiments only to find that the resulting data proves difficult to analyze with traditional visualization techniques. A typical laboratory experiment, for example, will setup several systems in each of two or more groups-a control group and various treatment groups. The investigator will then measure various parameters for each system over time (sometimes dozens to hundreds of parameters). The investigator tries to answer this question: What is different between the control and treatment groups? Field monitoring of polluted and unpolluted sites within a region result in the same kinds of data and the same difficulties in visualization. Plotting multiple lines on a single 2D plot quickly gets confusing. Plus, the dynamic interactions between parameters can be hard to see. 3D visualizations help researchers make qualitative insights about data more easily. We developed a 3D plotting technique, called “worm plots”, for visualizing these kinds of data. Our visualization tool examines how groups of points change over time. If we plot circles for each time slice, then these circles will change their relative sizes and positions as time goes by. To visualize these evolving dynamics, we connect these time slices with conic sections. This gives us spacetime worms that can be displayed graphically. In our experience, these simple circular cross sections work best when exploring data. They can be rendered quickly and give an easily intuited summary of each group. Further enhancements tend to make the image busy and difficult to interpret     

Machine learning to boost the next generation of visualization technology.

Visualization has become an indispensable tool in many areas of science and engineering. In particular, the advances made in the field of visualization over the past 20 years have turned visualization from a presentation tool to a discovery tool. Machine learning has received great success in both data mining and computer graphics; surprisingly, the study of systematic ways to employ machine learning in making visualization is meager. Like human learning, we can make a computer program learn from previous input data to optimize its performance on processing new data. In the context of visualization, the use of machine learning can potentially free us from manually sifting through all the data. This paper describes intelligent visualization designs for three different applications: (1) volume classification and visualization, (2) 4D flow feature extraction and tracking, (3) network scan characterization.     

Fast, memory-efficient cell location in unstructured grids for visualization

Applying certain visualization techniques to datasets described on unstructured grids requires the interpolation of variables of interest at arbitrary locations within the dataset's domain of definition. Typical solutions to the problem of finding the grid element enclosing a given interpolation point make use of a variety of spatial subdivision schemes. However, existing solutions are memory- intensive, do not scale well to large grids, or do not work reliably on grids describing complex geometries. In this paper, we propose a data structure and associated construction algorithm for fast cell location in unstructured grids, and apply it to the interpolation problem. Based on the concept of bounding interval hierarchies, the proposed approach is memory-efficient, fast and numerically robust. We examine the performance characteristics of the proposed approach and compare it to existing approaches using a number of benchmark problems related to vector field visualization. Furthermore, we demonstrate that our approach can successfully accommodate large datasets, and discuss application to visualization on both CPUs and GPUs.