Blog
Thoughts on AI, product development, and building in the open.
The Morphological Structure of English Words: A Corpus-Based Analysis of 7,533 Word Families
We constructed and analyzed a large-scale derivation tree corpus covering 7,533 word families and 54,795 derivation nodes. Latin and Greek roots account for 50.8% of all families, and 86.4% of word families can be decoded within just two affixation steps.
FSRS-5.0 vs Traditional Spaced Repetition: A Data-Driven Comparison
We compared four spaced repetition algorithms over 180-day simulations. FSRS-5.0 achieves 28% fewer reviews than SM-2 while maintaining comparable retention, yielding a 37% improvement in learning efficiency.