Process Oriented Discourse Analysis - Technologies for Discourse-Based Research in Media History and the History of Science

该项目旨在建立一个虚拟的研究环境，以便在人文学科，特别是历史学科中进行话语分析。在包豪斯-魏玛大学和雷根斯堡大学的媒体学者、计算机科学家和图书馆员的合作基础上，鉴于19世纪和20世纪初媒体技术与实验室科学之间的关键关系，将开发和建立这种研究环境。设想的虚拟研究环境将(A)为历史话语分析中的技术驱动项目提供一个理论上扎根的程序框架，(B)提供量身定制的用户界面和工具，用于有效地分析和建模相应的研究问题，以及(C)在魏玛和雷根斯堡的大学图书馆实施，供其他学者不受限制地使用。与现有的研究环境相比，我们的方法将更容易将用户的专业知识动态整合到分析方法和工具中。潜在的文本分析管道将基于用户针对其研究主题设计的显式语义模型，部分利用主动机器学习。自然语言处理、分析算法和基于知识的文本处理将被整合，以应对复杂的语篇分析过程。在这方面，用户将能够通过专门设计的话语查询语言使用这些工具。我们计划的研究环境还将提供范例功能，如‘话语地图’和‘话语透镜’，以便连接虚拟研究环境和由此衍生的学术出版物。我们的项目将针对1860年生命科学中的所谓‘实验室革命’而开发，指的是并依赖于媒体研究和科学史中处理这一主题的广泛学术知识。利用现有的‘虚拟实验室’语料库，并根据目前的项目进一步扩展，我们的研究环境将在其发展的每个阶段进行测试和评估。总的目标是模拟和绘制“实验室革命”的所有话语参与者(即科学家、工程师、仪器、生物体、概念、地点等)。关于他们在将实验室的过程和程序转化为媒体技术方面所做的贡献，反之亦然。

This project aims at establishing a virtual research environment for conducting discourse analyses in the humanities, in particular history. Based on a cooperation between media scholars, computer scientists, and librarians located at the Bauhaus-UniversitätWeimar and the University of Regensburg, this research environment will be developed and established in view of the crucial relation between media technologies and laboratory science in the 19th and early 20th centuries. Once established, it can be used in the context of similar projects carriedout in neighboring disciplines.The envisaged virtual research environment will (a) offer a theoretically grounded procedural framework for technology-driven projects in historical discourse analyses, (b) provide tailored user interfaces and tools for efficiently analyzing and modeling the corresponding research questions, and (c) be implemented at the University libraries in Weimar and Regensburg for unrestricted use by other scholars.In contrast to existing research environments our approach will ease the dynamic integration of the expertise of its users into the analysis methods and tools. The underlying text analytics pipeline will be grounded on explicit semantic models that users devise with regard to their research topics, partly drawing on active machine learning. Natural language processing, analysis algorithms, and knowledge-based text processing will be integrated to cope with the complex process of discourse analysis. In this regard, users will be able to use These tools by means of a specifically designed Discourse Query Language. Our planned research Environment shall also provide paradigmatic features such as 'Discourse Maps' and 'Discourse Lenses' in order to connect the virtual research environment and scholarly publications derived from it.Our project will be developed with respect to the so-called 'laboratory revolution' in the life sciences in 1860, referring to and relying on the extensive scholarship treating this topic in media studies and the history of science. Using the already existing corpus of the 'Virtual Laboratory' and further expanding it in view of the present project, our research environment will be tested and evaluated at every stage of its development. Overall goal is to model and map all discourse actors of the 'laboratory revolution' (i.e. scientists, engineers, instruments, organisms, concepts, sites, etc.) with regard to their contribution to translating and transposing processes and procedures from the laboratory into media technologies and vice versa.