Word2Vec: Optimal hyperparameters and their impact on natural language processing downstream tasks
Word2Vec is a prominent model for natural language processing tasks. Similar inspiration is found in distributed embeddings (word-vectors) in recent state-of-the-art deep neural networks. However, wrong combination of hyperparameters can produce embeddings with poor quality. The objective of this wo...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
De Gruyter
2022-03-01
|
| Series: | Open Computer Science |
| Subjects: | |
| Online Access: | https://doi.org/10.1515/comp-2022-0236 |