Automatic Dialect Adaptation in Finnish and its Effect on Perceived Creativity
This work addresses the challenge of dialect adaptation in Finnish for applications like creative text generation, though it is incremental as it builds on existing NMT methods.
The authors tackled the problem of automatically adapting standard Finnish text to various dialects using character-level NMT models, finding that transfer learning slightly outperformed a multi-dialectal approach across over 20 dialects. They also studied how dialect adaptation affects perceived creativity in computer-generated poetry, revealing that greater deviation from standard Finnish led to lower scores on an existing metric, but people associated creativity more with dialects and fluency more with standard Finnish.
We present a novel approach for adapting text written in standard Finnish to different dialects. We experiment with character level NMT models both by using a multi-dialectal and transfer learning approaches. The models are tested with over 20 different dialects. The results seem to favor transfer learning, although not strongly over the multi-dialectal approach. We study the influence dialectal adaptation has on perceived creativity of computer generated poetry. Our results suggest that the more the dialect deviates from the standard Finnish, the lower scores people tend to give on an existing evaluation metric. However, on a word association test, people associate creativity and originality more with dialect and fluency more with standard Finnish.