VIS x GenAI 2025

The "Yearly Downloads Trend" chart reveals a dynamic growth pattern in downloads from 1990 to 2024, with notable fluctuations. Downloads steadily increased from 1990 to 2007, peaking sharply between 2005 and 2007. The highest download activity occurred in 2010, reaching 156,545, followed by a gradual decline after 2011. A significant drop is observed post-2020, with downloads plummeting to just 31,585 in 2024, indicating a possible shift in user behavior or market dynamics. The data suggests a strong growth phase in the early 2000s, followed by stabilization and eventual decline, warranting further investigation into external factors influencing these trends.

The chart titled "Conference Impact on Downloads" reveals significant disparities in download counts across four conferences: InfoVis, VAST, Vis, and SciVis. InfoVis leads with a remarkable 1,278,342 downloads, far surpassing the other conferences, indicating its dominant influence and popularity in driving content engagement. VAST and Vis follow with 889,239 and 852,490 downloads, respectively, showcasing strong but relatively comparable impacts. SciVis, however, lags considerably with only 269,424 downloads, suggesting either lower visibility or niche appeal. This data highlights InfoVis as the most impactful conference in terms of download activity, while SciVis may benefit from strategies to enhance its reach and engagement.

The chart reveals a clear upward trend in download activity from 1990 to a peak in 2010, followed by a gradual decline with sharper drops after 2020. Notably, downloads exhibit significant year-to-year fluctuations, suggesting seasonal or cyclical variations. The sharp growth between 2000 and 2010 indicates a period of rapid adoption or increased demand, potentially tied to technological advancements or market expansion. However, the steep decline post-2020 may reflect market saturation, shifts in user behavior, or external disruptions. The data highlights the importance of understanding both long-term trends and short-term variability in download patterns.

The scatter plot reveals a positive correlation between downloads and citation counts across both Aminer and CrossRef datasets, though the strength of the relationship varies. Higher download counts generally correspond to higher citation counts, particularly for Aminer Citation Count, which exhibits a wider spread at higher download values. CrossRef Citation Count displays a more concentrated distribution, with many papers having zero citations despite moderate download activity. This suggests that while downloads may indicate interest, citation impact depends on additional factors such as research quality or relevance. The opacity and size of markers effectively highlight overlapping data points, emphasizing the density of low-citation papers.

The chart reveals distinct citation trends for Aminer and CrossRef over time, with both datasets showing peaks and declines. Aminer citation counts exhibit a sharp rise from 1990 to 2007, peaking in 2007 at 14,757 citations, followed by a steady decline after 2010, with near-zero citations by 2024. In contrast, CrossRef citations show a slower but consistent growth until 2007, peaking at 7,210 citations, and then experiencing fluctuations with a gradual decline, also nearing zero by 2024. The divergence in citation trajectories suggests varying relevance or data availability for these sources over time, with Aminer showing a more dramatic drop-off compared to CrossRef. This could reflect shifts in academic citation practices or database coverage.

The chart reveals significant disparities in citation counts across conferences, with "Vis" consistently leading in both Aminer and CrossRef citation counts (120,100 and 37,925, respectively). "InfoVis" follows closely in Aminer citations (102,736) but surpasses "Vis" in CrossRef citations (50,758). "VAST" and "SciVis" trail significantly, with "SciVis" showing the lowest citation counts in both datasets. The discrepancy between Aminer and CrossRef citation counts is notable, particularly for "Vis" and "InfoVis," suggesting potential differences in citation coverage or indexing practices between the two sources.

The chart reveals a clear upward trend in award counts from 1990 to 2024, with notable periods of acceleration and stabilization. From 1990 to the early 2000s, awards were relatively sparse, fluctuating between 1 and 5 annually. However, starting in the mid-2000s, there was a marked increase, with awards peaking sharply in 2010 at 15, followed by a slight dip and recovery. The most striking growth occurred from 2020 onward, where awards surged to a consistent high of 20 in 2021 and 2022, before slightly leveling off at 19 in 2023 and 2024. This suggests a significant rise in recognition or output in recent years, potentially driven by industry expansion, increased competition, or evolving criteria for awards.

The scatter plot reveals a notable disparity between awards and citation counts, with a wide range of citation counts observed across different award categories. While some award categories (e.g., 'TT') are associated with consistently high citation counts, others, such as 'Unknown,' display significantly lower citation counts, suggesting a potential lack of correlation between recognition through awards and academic impact as measured by citations. The clustering of data points for certain award types indicates that specific awards may be tied to higher research visibility or productivity, while others may not. This highlights the variability in how awards align with citation-based metrics of scholarly influence.

The box plot reveals significant disparities in citation counts across award categories. The 'TT' group exhibits the highest citation counts, with a wide range and extreme outliers, indicating that recipients of this award tend to have substantial academic influence. The 'BP' and 'HM' groups show moderate citation counts with a broader spread, suggesting variability in their recipients' impact. Conversely, 'BA' and 'BCS' have lower citation counts and narrower distributions, indicating a more consistent but less impactful citation profile. The 'TT;BP' group, a hybrid category, shows a small sample size but includes a mix of high and low citations, reflecting diverse academic recognition. Overall, the chart highlights that awards like 'TT' are associated with higher academic influence, while others like 'BA' and 'BCS' are linked to more modest citation profiles.