Visual Text Analytics

视觉文本分析

• 批准号: 1000228345-2012
• 负责人: Matwin, Stan
• 金额: $ 10.93万
• 依托单位: Dalhousie University
• 依托单位国家: 加拿大
• 项目类别: Canada Research Chairs
• 财政年份: 2019
• 资助国家: 加拿大
• 起止时间: 2019-01-01 至 2020-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=693842
• 关键词: Visual; Text; Analytics

The Big Data problem is recognized as one of the global, strategic challenges facing the developed societies of today. Mining, understanding, and using text data is at the centre of Big Data: text is ubiquitous, and its free structure presents challenges far beyond those existing for numerical data analytics. This proposal recognizes that Visual Text Analytics (VTA) is the emerging tool to address Big Text Data. VTA proceeds in a cycle: Analyze first, Show the Important, Zoom, Filter, and Analyze further. The proposed research program will address all the elements of this cycle. We will investigate new linguistically-informed text models, representations and transformations, new text analytics techniques (particularly as the data grows in a never-ending stream), and novel ways to visualize the results of text analytics. Emphasis will be on solutions that scale-up and are verified in real-life, industrially-relevant applications.

大数据问题被认为是当今发达社会面临的全球性战略挑战之一。挖掘、理解和使用文本数据是大数据的核心：文本无处不在，其自由结构带来的挑战远远超出了现有的数字数据分析。本提案认识到可视化文本分析（VTA）是解决大文本数据的新兴工具。VTA按一个循环进行：首先分析，显示重要内容，放大，筛选，进一步分析。拟议的研究计划将涉及这一周期的所有要素。我们将研究新的基于语言的文本模型、表示和转换、新的文本分析技术（特别是当数据以永无止境的流增长时），以及将文本分析结果可视化的新方法。重点将放在可扩大规模并在实际工业应用中得到验证的解决方案上。