End-User Programming of Visualizations

可视化已经成为传达信息的重要手段。随着越来越多没有编程和数据分析背景的人可以访问数据，对更多以用户为中心的可视化软件的需求日益增长。本研究计划旨在利用程式设计的注意力投资模型，从终端用户发展的角度来研究可视化创作。信息可视化已经成为许多人的共同活动。可视化的范围从简单的图表到信息图表，是由记者发布他们的发现或设计师以富有表现力的、引人注目的可视化方式覆盖整个主题而创建的，可以针对新闻读者、商业人士或研究受众。考虑到使用可视化作为交流手段的人群非常不同，创建这些可视化的方法也多种多样。Victor（2013）将这些方法分为三种范式：使用，绘图和编码，每种都体现在不同的软件应用程序中。代表性的例子分别是Microsoft Excel， Adobe Illustrator和d3 .js。本研究计划建议从终端用户编程的角度来看待数据可视化。从最终用户的角度来看，学习一门用于创建可视化的编程语言，比如Processing或Data-Driven Documents (D3.js) (Bostock et al.2011)，太昂贵了。因此，终端用户更喜欢使用更直接的信息操作，例如电子表格（Lieberman et al . 2006）。然而，关于信息可视化主题的研究似乎很少以用户为中心的方式进行，即使它是针对最终用户的（Pantazos et al. 2013）。为可视化的最终用户编程提出的方法也借鉴了注意力投资理论（Blackwell 2002, Blackwell & Green 1999）。该理论已经成功地应用于各种电子表格功能（Beckwith et al.2002, Jones et al. 2003），但没有应用于数据可视化。在注意力投资模型的指导下，目的是研究信息可视化创建的认知方面，现有系统如何支持用户并为数据可视化创建原型，这些原型考虑了最终用户的感知成本、风险和收益。

Visualizations have become important means of communicating information. As databecomes accessible for more people with no background in programming and dataanalytics, there is a growing need for more user-centered visualization software. Thisresearch proposal aims to investigate visualization creation from an end user developmentperspective, using the Attention Investment model of programming.Information visualization has become a common activity for many people. Visualizationsrange from simple graphs to infographics, are created by journalists publishing theirfindings or designers covering an entire topic in an expressive, compelling visualization,and can target news readers, business people or research audiences. Given the verydifferent groups of people that are using visualizations as means of communication, theways of creating these visualizations are diverse. Victor (2013) classifies these ways inthree paradigms: using, drawing and coding, each embodied in different softwareapplications. Representative examples would be Microsoft Excel, Adobe Illustrator andD3.js respectively.This research project proposes to look at data visualization from the perspective of enduser programming. From the end user perspective, learning a programming language forcreating visualizations, like Processing or Data-Driven Documents (D3.js) (Bostock et al.2011), is too expensive. As such, end users prefer to use a more direct manipulation ofinformation, for example spreadsheets (Lieberman et al 2006). However, research on thetopic of information visualization very rarely appears to be done in a user-centric manner,even if it's directed for end users (Pantazos et al. 2013).The approach proposed for end user programming of visualizations also draws on thetheory of Attention Investment (Blackwell 2002, Blackwell & Green 1999). The theory hasbeen successfully applied mostly to various spreadsheet functionalities (Beckwith et al.2002, Jones et al 2003), but not to data visualizations.Guided by the Attention Investment model, the aim is to study the cognitive aspects ofinformation visualization creation, how existing systems support the users and createprototypes for data visualization that accounts for the perceived costs, risks and benefits ofthe end users.