基于数据驱动和图形语法的数据可视化引擎的研究与实现

发布时间：2018-03-06 03:16

本文选题：数据可视化　切入点：数据驱动　出处：《浙江大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着大数据时代的到来,如何采用有效的方法对数据进行更好的处理分析变的越来越重要,而数据可视化是数据分析中一种极为重要的方法,它不仅能够直观的展现数据之间的关系,更重要的是能够帮助数据分析师和管理者迅速的做出重要决策。本文主要的贡献有:(1)本文提出了一种基于数据驱动和图形语法的数据可视化图表引擎,专门用于数据到统计图形的可视化表达。相比于一般的可视化图形库,本文所述的数据可视化引擎提供了一种基于图形语法的、用户友好的方法来用于描述图形结构,并且提供声明式的中间层语义描述来更细粒度的描述图表。通过这种方法可以组合出更多类型的图表。(2)本文基于HTML5的Canvas开发了一套轻量快速的底层渲染库并且提供了数据驱动的选择集合机制对进行场景树进行更加高效的管理。(3)本文通过对数据分面来实现高维多元数据的透视划分,并提出一种图形透视表的进行透视数据可视化,有效的帮助理解高维多元数据。本文提出的技术以及实现的引擎已经应用在网易互联网多款产品中,同时为敏捷分析平台网易有数提供可视化图表渲染,为数据分析和决策提供了快速的可视化帮助。
[Abstract]:With the arrival of big data era, how to use effective methods to process and analyze data becomes more and more important, and data visualization is one of the most important methods in data analysis. It can not only show the relationship between the data intuitively, More importantly, it can help data analysts and managers to make important decisions quickly. Compared with the general visual graphics library, the data visualization engine described in this paper provides a graphical syntax based, user friendly method to describe the graphic structure. And provide declarative middle layer semantic description to finer grained description diagram. This method can combine more types of charts. 2) this paper develops a set of lightweight and fast bottom rendering library based on HTML5 Canvas. This paper provides a data-driven selection set mechanism to manage the scene tree more efficiently. 3) in this paper, the perspective partition of high-dimensional multi-metadata is realized by dividing the data. And a graphic PivotTable for perspective data visualization, effective help to understand high-dimensional multivariate data. The technology proposed in this paper and the implementation of the engine has been used in NetEase Internet products, At the same time, it provides visual chart rendering for NetEase, an agile analysis platform, and provides a quick visual aid for data analysis and decision making.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41

【相似文献】