Morphological Typology in BPE Subword Productivity and Language Modeling
This addresses the problem of optimizing tokenization for language modeling across different languages, though it appears incremental as it builds on existing BPE methods and typological theories.
The study tackled the problem of how morphological typology affects tokenization and language modeling, finding that languages with synthetic features show greater subword regularity and productivity with BPE tokenization, leading to better language modeling results.
This study investigates the impact of morphological typology on tokenization and language modeling performance. We focus on languages with synthetic and analytical morphological structures and examine their productivity when tokenized using the byte-pair encoding (BPE) algorithm. We compare the performance of models trained with similar amounts of data in different languages. Our experiments reveal that languages with synthetic features exhibit greater subword regularity and productivity with BPE tokenization and achieve better results in language modeling tasks. We also observe that the typological continuum from linguistic theory is reflected in several experiments. These findings suggest a correlation between morphological typology and BPE tokenization efficiency.