A light-weight demo app for geospatial semantic search. Designed for any kind of textual data with geospatial references. Building on previous research:
- An Application-Oriented Implementation of Hexagonal On-the-fly Binning Metrics for City-Scale Georeferenced Social Media Data
- Developing a Privacy-Aware Map-Based Cross-Platform Social Media Dashboard for Municipal Decision-Making
Repository for the paper: XXX (still to submit)
The paper describes an approach to use semantic similarity for geospatial purposes, like georeferenced social media data.
- 10.000 random individual posts (10Mb)
- 20.000 random individual posts (20Mb)
- 40.000 random individual posts (40Mb)
- 693.959 random aggregated posts, mean embedding for 6630 locations (13Mb)
- 693.293 random aggregated posts, mean embedding for 5964 locations with >1 posts (12Mb)
- 686.558 random aggregated posts, mean embedding for 4487 locations with >9 posts (9Mb)
- 542.722 random aggregated posts, mean embedding for 1028 locations with >99 posts (2Mb)