Abstract

This paper explores the contribution of software-based tools that are increasingly used for the semi-automated analysis of large volumes of text, especially Topic Modelling and Corpus Linguistics. These tools highlight the potential of getting interesting and new insights quickly, but at a cost. Linguistic aspects need to be considered carefully if computer-assisted technologies are to provide valid and reliable results. Main features of these tools will be presented, and some general problems and limitations will be discussed. The relation between technical tools and theoretical frameworks is discussed. The main empirical reference is the case of climate change.

摘要

本文探讨了基于软件的工具的贡献，这些工具越来越多地用于对大量文本进行半自动分析，尤其是主题建模和语料库语言学。这些工具突出了快速获得有趣和新见解的潜力，但需要付出代价。如果计算机辅助技术要提供有效和可靠的结果，就需要仔细考虑语言方面的问题。将介绍这些工具的主要特点，并讨论一些一般问题和限制。讨论了技术工具和理论框架之间的关系。主要的经验参考是气候变化的案例。