← Names and Name-Based Inference

We compare name classification using phonetic encodings with classification using character sequences. In the reported experiments, an LSTM trained on Metaphone encodings performs worse, suggesting that spellings contain information lost by the sound encoding.

Cite: Gaurav Sood. "Sound Names: Classify Names Based on Sequence of Sounds." https://gsood.com/research/sound-names-classify-names-based-on-sequence-of-sounds/

BibTeX:
@misc{soodsound,
  author = {Gaurav Sood},
  title = {{Sound Names: Classify Names Based on Sequence of Sounds}},
  url = {https://github.com/appeler/sound_names}
}