CiteSpace 7.0 presents several major changes since 6.4.R2, including an optimized performance with much larger networks than previous versions, upgraded cluster summarizations with more recent large-language models, more streamlined save and reopen saved visualization, and fixed unstable functions such as Save PNG and rotating labels in timeline views. The current 7.0 is still an experiment version as a few key features are still under constructions due to the substantial internal optimization. This blog will serve as a living document to share with you the latest available features of the new version 7.0.

Since its first publication in PNAS in 2004, CiteSpace has evolved over more than two decades as a visual analytic tool designed to reveal macroscopic dynamic patterns and evolutionary trends in scientific literature. On Christmas Day 2025, I searched in several major scholarly databases for publications related to CiteSpace, and I am pleased to share the snapshot of numbers below, together with selected academic fields and research themes in which CiteSpace has been applied.

A co-citation network is a rich source of inspirational insights, intriguing connections, maybe even transformative surprises. A large enough co-citation network is necessarily complex structurally, temporarily, and above all, intellectually. Overtime, layers, patches, as well as new branches of networks would emerge and morph into existing ones. What can we learn from a co-citation network? How would our newly learned lessons alter our understanding of how a research field finds its own way ahead with competing as well as encouraging lights shed from a diverse spectrum of independent and creative minds?

In this blog, I illustrate how models such as gpt-4o, gpt-4o-mini, and deepseek-chat can augment our sensemaking of clusters of co-cited intellectual contributions. A practical question is: how can we determine systematically and objectively which cluster summaries are the most reliable choices? Furthermore, from a sustainability perspective: what factors should guide our decision when weighing affordable models against the most powerful ones?

Here are 9 scientometric and visual analytic reviews of various aspects of mind-body health. Using CiteSpace, each of them depicts thematic trends and concentrations based on tens of thousands of relevant publications.

Visualizing real-world networks often presents two intertwined challenges: (1) the visualization must reveal the distinctive structural and temporal characteristics of complex data, and (2) it must do so with a sufficient degree of clarity to make it meaningful. Here are some steps to optimize network layout in CiteSpace for users to focus on extracting valuable insights from the evolving patterns embedded in dynamic networks.

Here are some highlights based on 6,778 publications on the Web of Science Core Collection between 2004 and July 7, 2025. On the one hand, it shows some interesting developments concerning the contexts in which CiteSpace has been used. On the other hand, it illustrates the types of patterns and trends one may discern from such visual analytic studies.

By default, users create a new project in CiteSpace through an interactive process with the GUI. It is straightforward, but in rare occasions, an alternative approach could be handy. For example, if the browse button is not working for some unknown reason, you can bypass it by directly editing the project list file. Also, if you need to create multiple projects together, this alternative approach could be more efficient.

Scopus is a major source of data for bibliometric and scientometric studies. I will demonstrate the key steps of preparing your dataset with Scopus and visualizing trends and patterns in scientific literature with the current version of CiteSpace 6.4.R2.

The CiteSpace Q&A GPT, formerly called CiteSpace 101 GPT, is created to help you with conceptual and practical questions about CiteSpace.

CiteSpace’s latest enhancement with GPT-4o ensures distinct cluster labels and high-quality citation-integrated context-aware summaries, bringing AI-assisted reviews closer to expert-level analysis. CiteSpace 最新增强功能结合 GPT-4o，确保生成独特的聚类标签和高质量的引文整合型上下文感知摘要，使 AI 辅助文献综述更接近专家级分析。

Researchers pursue new ideas quite often if they don't do that all the time. A commonly known approach to be creative is to connect previously disparate concepts, topics, opinions, and the like. Sometimes more and more people are convinced and they start to follow, expand, or extend the newly conceived connection. Other times, reactions may be limited. In a 2012 JASIST paper (Chen 2012), I initially proposed the Structural Variation Theory to guide the search for new publications with a great potential to change the landscape of a field of research.