~/talk

Beagle: Automated Extraction and Interpretation of Visualizations from the Web

on YouTube 19m
Talkat CHI 2018 · 2018-04-21..04-26· [data-visualization]

speakers

name:Leilani Battle
links:none on file
name:Peitong Duan
links:none on file
name:Zachery Miranda
links:none on file
name:Dana Mukusheva
links:none on file
name:Remco Chang
links:none on file
name:Michael Stonebraker
links:none on file

resources

video:https://youtu.be/L0y59TJovHs
slides:none on file

summary

Beagle automatically crawls the web for SVG visualizations, extracting about 41,000 visualizations from five repositories using an SVG island crawl strategy, computes 100+ SVG-derived features, and trains a random forest that reaches roughly 85% accuracy across 24 types. Results show four types dominate (bar, line, scatter, map) — bar line scatter map dominance — pie charts under 5%, and the pipeline scales enough to build a queryable corpus for further analysis.

AI-written summary · about

chapters (11)

  1. 0:00Introduction and background
  2. 2:14Project contributions overview
  3. 4:22Web crawling challenges
  4. 6:30Island search technique
  5. 7:54SVG statistics and labeling
  6. 10:45Machine learning setup
  7. 12:13Annotated dataset analysis
  8. 13:39Key takeaways and insights
  9. 14:43Future work directions
  10. 15:26Summary and data availability
  11. 15:57Q&A session

related talks

~/search
type to search 6,249 conferences, 138,987 speakers and 185,172 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.