Interactive web application for exploring the temporal, causal, and character-based relationships of the television series Dark (Netflix, 2017--2020). Transforms flat CSV data into a typed Knowledge Graph and provides seven coordinated visualization views with a clean glassmorphic UI.
The main view. A force-directed graph with 4 node types: Event (circles, colored by world), Character (diamonds, sized by appearance), World (hexagons), and TimePeriod (pills). 8 edge types connect them. The left sidebar contains filters, toggles, legend, and the entity inspector panel.
Swimlane-based timeline where events are grouped into time-boxes (clustered by year proximity). Smooth Bezier curves show causal transitions between events across adjacent time windows.
Event-centric force layout. Events are linked when they share one or more characters. Node size encodes importance (number of connected events). Filterable by world and event type.
Grouped vertical bars with configurable grouping (by year, world, or top characters) and selectable metrics (event count, death events, important triggers). Animated D3 transitions on filter change.
Summary dashboard showing 7 key stat cards (total events, characters, worlds, deaths, triggers, etc.) plus 4 chart types: events-by-year bar, edge-type donut, top-20 character involvement bar, and events-by-world distribution pie.
Character co-occurrence chord diagram. The top 14 most frequent characters are shown as arc segments around the circle. Ribbons between arcs represent events where both characters appear — thicker ribbons mean more co-occurrences. Hover over an arc to see total co-occurrences, or hover a ribbon for the specific pair count.
Beeswarm-per-lane layout across World bands covering 1885--2056. Each row is a world (Jonas, Martha, Origin). Dots are events — click any dot to see details in the bottom detail panel. Filterable by character, event type, and world.
- Serve the project directory:
python -m http.server 8000 - Open
http://localhost:8000in a browser. - A local server is required;
file://will not work due to CORS restrictions on XHR-based CSV loading.
Abdullah Al Mamun
M.Sc. & B.Sc. in Software Engineering
TU Wien (Vienna, Austria) & Daffodil International University
Email: mamun.swe.de@gmail.com
GitHub: github.com/abbysweb
ORCID: 0009-0006-7473-0024