Bat Sonotype Classification Data from Hong Kong Field Recordings

Hong Kong field recordings of bats, with extracted call parameters and supporting R code for classification. The dataset is 88.9 MB in size, authored by John Allcock, and was last updated on April 13, 2026. It is shared under a CC-BY-4.0 license.

Use Cases

• Train supervised classification models for bat species based on extracted call parameters.
• Develop unsupervised clustering methods for defining sonotypes in regions lacking reference libraries.
• Benchmark acoustic identification algorithms using real-world field recordings from a tropical landscape.
• Analyze the relationship between bat call parameters and species or environmental factors.

Strengths

• Includes both extracted call parameter data and the R code for the classification methodology.
• Dataset size is 88.9 MB, indicating a substantive collection of audio features and code.
• Released under a permissive CC-BY-4.0 license, facilitating reuse and adaptation.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for large-scale model training.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

figshare

Collection Method

Call parameters extracted from field recordings of bats.

Time Range

null

Freshness

Last updated 2026-04-13 09:54:36; freshness should be verified.

Geography

Hong Kong