A Study on Improving Semantic Consistency of Translation Systems by Combining Dynamic Computing Methods in English Corpus
Bibliographic record
Abstract
Aiming at the dilemma of corpus-based intelligent English translation, the article proposes an English neural machine translation method based on depth-separable convolution, which combines with the dynamic computation method to improve the semantic consistency of the translation system for semantic alignment and fusion.In order to verify the training effect of the proposed convolutional neural network model combined with the dynamic computation method, comparison experiments with one-way and two-way network models and baseline model with different cut-off granularity are conducted respectively.In order to better examine its performance in practical translation applications, online translation, machine translation and systematic methods are utilized for comparison.The BLUE values of this paper's model for Chinese-English data translation in four different granularities of words, syllables, subwords and characters are 21.41%, 21.91, 29.25% and 20.40%, respectively.In 100,000, 200,000 and 500,000 training English-Chinese bilingual parallel corpus, the training time consumed by the model in this paper is 9.58 h, 15.94 h and 32.69 h.In practical application, the decibel range of the noise reduction of the translation system method designed by the research is distributed in [1.62 ~ 1.89], the average value of coherence is 91.1%, and the average compression rate and the average stability of the BLEU scores are 93.84% and 98.38%, respectively, and the results are better than the comparison methods.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.008 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.004 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".