Scalable Visual Analysis of Patent and Scientific Document Collections

专利和科学文档集的可扩展可视化分析

• 批准号: 81590254
• 负责人: Professor Dr. Thomas Ertl
• 金额: --
• 依托单位: Centrum für Informations- und Sprachverarbeitung (CIS)
• 依托单位国家: 德国
• 项目类别: Priority Programmes
• 财政年份: 2008
• 资助国家: 德国
• 起止时间: 2007-12-31 至 2017-12-31
• 项目状态: 已结题
• 来源: https://gepris.dfg.de/gepris/projekt/81590254?language=en
• 关键词: Scalable; Visual; Analysis; Patent; Scientific

In addition to patent documents, other information sources, such as scientific literature, play an important role in many tasks in intellectual property (IP) analysis. Particularly the detection of shifts in scientific research, of emerging trends, and of promising technological innovations, which we will refer to as white spot analysis, are of great interest to different knowledge workers. These tasks, however, are almost impossible to perform in a fully automated way, since interpretation of the results requires background knowledge and the experience of human users. The main goal of this proposal is, therefore, to develop a new visual analytics approach for white spot detection by combining sophisticated text analysis techniques and interactive visual methods to facilitate sensemaking in the field of scientific literature. Subgoals are the identification of topic shifts and changes in citation networks over time, including provenance tracking to help users understand uncertainties of their analysis and to support collaborative scenarios on high resolution displays as well as web-based interfaces. These will leverage proactive, visual guidance to potentially interesting findings (white spots), thereby relieving users of the burden to explore large, complex, and high-dimensional document spaces manually. By combining the knowledge of both groups in text analysis, interactive visualization, and the experiences gained during the previous funding period, this project aims to develop support systems for strategic tasks in the intellectual property domain that improve the current state of the art regarding topic and citation analysis as well as white spot identification.

除了专利文件外，科学文献等其他信息来源在知识产权（IP）分析的许多任务中也发挥着重要作用。特别是对科学研究中的变化、新兴趋势和有希望的技术创新的检测，我们将其称为白色点分析，这对不同的知识工作者都很有兴趣。然而，这些任务几乎不可能以完全自动化的方式执行，因为结果的解释需要背景知识和人类用户的经验。因此，本提案的主要目标是开发一种新的视觉分析方法，通过结合复杂的文本分析技术和交互式视觉方法，以促进科学文献领域的意义构建，用于白色斑点检测。子目标是识别引文网络随时间的主题转移和变化，包括出处跟踪，以帮助用户了解其分析的不确定性，并支持高分辨率显示器和基于Web的界面上的协作场景。这些将利用前瞻性的视觉指导来发现潜在的有趣发现（白色点），从而减轻用户手动探索大型、复杂和高维文档空间的负担。通过结合两个小组在文本分析、交互式可视化方面的知识以及在上一个资助期获得的经验，该项目旨在为知识产权领域的战略任务开发支持系统，以改进主题和引文分析以及白色点识别方面的当前技术水平。