Created in the context of the 15th IEEE Pacific Visualization Symposium, 11.-14. April 2022 (Online).

A scientific article can be cited with different intents over several years. The citation intents can be inferred by classifying the citation text into different categories. With multiple citations to the same article, the citation intent categories overlap, making their analysis more challenging. We model the categories as dynamic sets and propose an approach to visualize temporal citation trends of an article across overlapping citation intents. The approach supports comparison between the citation trends of two seed articles of interest. The implemented prototype supports searching and selecting seed articles from a Semantic Scholar dataset.