Abstract Purpose We present an analytical, open source and flexible natural language processing and text mining method for topic evolution, emerging topic detection and research trend forecasting for all kinds of data-tagged text. Design/methodology/approach We make full use of the functions provided by the open source VOSviewer and Microsoft Office, including a thesaurus for data clean-up and a LOOKUP function for comparative analysis. Findings Through application and verification in the domain of perovskite solar cells research, this method proves to be effective. Research limitations A certain amount of manual data processing and a specific research domain background are required for better, more illustrative analysis results. Adequate time for analysis is also necessary. Practical implications We try to set up an easy, useful, and flexible interdisciplinary text analyzing procedure for researchers, especially those without solid computer programming skills or who cannot easily access complex software. This procedure can also serve as a wonderful example for teaching information literacy. Originality/value This text analysis approach has not been reported before.

中文翻译：

---

基于开源软件的话题演化与新兴话题分析

摘要目的我们提出一种分析，开放源代码和灵活的自然语言处理和文本挖掘方法，用于各种数据标记文本的主题演化，新兴主题检测和研究趋势预测。设计/方法/方法我们充分利用了开源VOSviewer和Microsoft Office提供的功能，包括用于数据清理的同义词库和用于比较分析的LOOKUP函数。结果通过在钙钛矿太阳能电池领域的应用和验证，该方法被证明是有效的。研究局限性为了获得更好，更具说明性的分析结果，需要一定数量的手动数据处理和特定的研究领域背景。也需要足够的分析时间。实际意义我们试图建立一个简单，有用，以及针对研究人员的灵活的跨学科文本分析程序，尤其是那些没有扎实的计算机编程技能或无法轻松访问复杂软件的人员。此过程也可以作为教学信息素养的一个很好的例子。原创性/价值以前没有报告过这种文本分析方法。