Achieving Human Parity on Automatic Chinese to English News Translation

[kweonwooj]

Abstract

• reports that the quality of Microsoft's Chinese-to-English machine translation on news sentences is at human parity
  • uses WMT2017 news translation data
  • defines how to accurately measure human parity in translation
  • describes the workflow and various experiments

Details

Defining Human Parity

• official definition : If there is no statistically significant difference between human quality scores for a test set of candidate translations from a machine translation system and the scores for the corresponding human translations then the machine has achieved human parity
• Evaluation method
  • use direct assessment described in WMT17
  • use source-based evaluation methodology described in IWSLT17
  • annotators are shown source text and a candidate translation and asked the questions How accurately does the above candidate text convey the semantics of the source text?, answering this using a slider ranging from 0 to 100 (100 being perfect)
  • to identify unreliable crowd workers, direct assessment includes artificially degraded translation output randomly

Neural Machine Translation

• LSTM, ConvS2S, Transformer are all SoTA models, but choose Transformer as baseline

Main Contributions

• Main Techniques used to achieve human parity
  • Careful data selection and filtering
  • Dual Learning to utilize the duality of the translation problem
  • Iterative joint training algorithm described in Zhang et al. 2018 to enhance the effect of monolingual source using Back Translation
  • Deliberation Network to refine translation based on two-pass decoding
  • New training objective over KL divergence to encourage agreement between left-to-right and right-to-left translation
  • System Combination and Re-ranking

Data Selection and Filtering

• Learn a bilingual sentence vector representation mapped into the same space to filter the noisy data and select relevant data
  • use method in Zoph et al. 2016 on subset of data known to be of good quality and relevant domain
  • use RNN enc-dec similar to GNMT as base model for representation learning
  • use cosine similarity of sentence representation of source S and target T
  • remove sentences with similarity below a specified threshold
• Rule-based filtering
  • both source and target sentence should contain at least 3 words, at most 70 words
  • pairs with ( src_len < 1.3 tgt_len, tgt_len < 1.3 src_len) removed
  • sentences with illegal chars (URL, char of other language) removed
  • Chinese sentence without any Chinese characters removed
  • duplicated sentence pairs are removed

Dual Learning

• Dual Unsupervised Learning use reconstruction log-likelihood of monolingual corpus while training
• Dual Supervised Learning trains primal and dual model simultaneously under regularization to encourage duality in probability distribution

Iterative Joint Training

Deliberation Network

L2R, R2L Agreement Regularization

• signals from R2L model can be leveraged to alleviate the exposure bias problem of L2R model and vice versa

System Combination and Re-ranking

• combine n-best hypotheses from all systems and train a re-ranker using k-best MIRA (margin-based classification algorithm)
• Features used for re-ranking are
  • original system score, 5-gram LM score, R2L score, Target2Source system re-score, cross-lingual sentence similarity between source and hypothesis
• turned out to be that original system score, LM score, R2L score, R2L sentence vector similarity and Target2Source sentence similarity were best features

NMT Pipeline

• Train ZhEn, EnZh Transformer model using DUL, DSL with bilingual corpus (multiple models can be trained to leverage ensemble)
• Generate back-translation corpus using En & Zh monolingual sentences and pre-trained models from previous step
• Train Transformer Model or Deliberation Network with inflated bilingual corpus, use pre-trained model's weight to initialize encoder and first-pass decoder of Deliberation Network

Experiments - Benchmark on WMT17

• Data
  • WMT17 EnZh 18M bilingual pairs. newsdev2017 as dev, newstest2017 as test set.
  • use LM trained on 18M bilingual pair to filter monolingual sentence from news.crawl and common.crawl
• Vocab
  • Byte Pair Encoding (BPE) of Zh 44k, En 33k
• Model
  • Transformer Big with Tensor2Tensor v1.3.0 open-source
  • 8x M40 GPUs
  • 200k Adam w learning_rate 0.3, decayed with noam schedule
  • 5120 words per batch, checkpoints created every 60 min
  • results are reported on averaged parameters of last 20 checkpoints
  • beam=8, length_penalty=1.0
  • reported score using sacreBLEU v1.2.3
• BLEU Score
  • Back-translation (BT) + Dual Learning + Deliberation Network combination performs best
  • Agreement Regularization does not add improvement much

Experiment on Larger Corpus

• Data
  • WMT17 18M + 35M/50M subset selected from 100M UN corpus
  • use Cross-Entropy selection by Moore et al. 2010 and Axelrod et al. 2011
  • use SentVect similarity filtering described in above Data Selection tab with threshold 0.2
• Vocab : same
• Model
  • Transformer Big with 8,192 hidden_size in conv-1 block (bigger than original Transformer Big)
  • 300K Adam
  • minibatch of 3,500 with 8 GPUs
  • same beam, length_penalty and averaging param
• BLEU Score
  • Base8k (larger model) performs better
  • additional corpus selected via SentVect enhances BLEU score best

Human Evaluation Results

• Ensembles (Combo-4,5,6) obtains human parity (equivalent score with Reference-HT)
  • Reference-HT are human translations without using online translation engines
  • Reference-PE are human post-edit output based on Google Translate results
  • Reference-WMT are original newstest2017 reference released after WMT17
  • Online-A-1710 : Microsoft Translator collected on Oct 2017
  • Online-A-1710 : Google Translator collected on Oct 2017

Evaluation Campaigns

• Motivated to resolve issues with human evaluation processes
  • Annotator variability : what if same annotator provides different results on same data? -> resolved by running three campaigns on same evaluation data, and has seen a near complete overlap
  • Data variability : conduct evaluations on completely different subsets of the test data. although, the test data may already be biased..

Human Analysis

• preliminary human error analysis on best system

Personal Thoughts

• Complete NMT workflow from data selection upto human evaluation and error analysis
• impressed on intense experiments, but methods are biased toward MS ideas
• human evaluation comparison between baseline model and improved model would have been interesting, see how the error types had been reduced

Link : https://www.microsoft.com/en-us/research/uploads/prod/2018/03/final-achieving-human.pdf Authors : Hassan et al. 2018