Credits

Kaiwu is built on research data and open resources. Here is what we use, and the citations their authors ask for.

Child language data (CHILDES)

Kaiwu marks the everyday words a child growing up in a Mandarin-speaking home hears most, so learners who grew up with Chinese meet them sound-first. We found them by counting what caregivers say in seven Mandarin corpora from CHILDES, part of TalkBank. Only word counts are used; the recordings and transcripts never leave our computer.

MacWhinney, B. (2000). The CHILDES Project: Tools for Analyzing Talk. 3rd ed. Mahwah, NJ: Lawrence Erlbaum.

Word frequency

To tell home words from words common everywhere, caregiver speech is compared with Chinese at large from wordfreq by Robyn Speer (data CC BY-SA 4.0).

Speer, R. (2022). rspeer/wordfreq: v3.0. Zenodo. doi:10.5281/zenodo.7199437

Stroke order

Stroke-order animations and tracing use data from Make Me a Hanzi and hanzi-writer-data, under the Arphic Public License.

Fonts

Inter, Noto Sans SC and Nunito, under the SIL Open Font License.