Alert button
Picture for Shaonan Zhang

Shaonan Zhang

Alert button

A neural interlingua for multilingual machine translation

Oct 16, 2018
Yichao Lu, Phillip Keung, Faisal Ladhak, Vikas Bhardwaj, Shaonan Zhang, Jason Sun

Figure 1 for A neural interlingua for multilingual machine translation
Figure 2 for A neural interlingua for multilingual machine translation
Figure 3 for A neural interlingua for multilingual machine translation
Figure 4 for A neural interlingua for multilingual machine translation

We incorporate an explicit neural interlingua into a multilingual encoder-decoder neural machine translation (NMT) architecture. We demonstrate that our model learns a language-independent representation by performing direct zero-shot translation (without using pivot translation), and by using the source sentence embeddings to create an English Yelp review classifier that, through the mediation of the neural interlingua, can also classify French and German reviews. Furthermore, we show that, despite using a smaller number of parameters than a pairwise collection of bilingual NMT models, our approach produces comparable BLEU scores for each language pair in WMT15.

* Accepted in WMT 18 
Viaarxiv icon

A practical approach to dialogue response generation in closed domains

Mar 28, 2017
Yichao Lu, Phillip Keung, Shaonan Zhang, Jason Sun, Vikas Bhardwaj

Figure 1 for A practical approach to dialogue response generation in closed domains
Figure 2 for A practical approach to dialogue response generation in closed domains
Figure 3 for A practical approach to dialogue response generation in closed domains
Figure 4 for A practical approach to dialogue response generation in closed domains

We describe a prototype dialogue response generation model for the customer service domain at Amazon. The model, which is trained in a weakly supervised fashion, measures the similarity between customer questions and agent answers using a dual encoder network, a Siamese-like neural network architecture. Answer templates are extracted from embeddings derived from past agent answers, without turn-by-turn annotations. Responses to customer inquiries are generated by selecting the best template from the final set of templates. We show that, in a closed domain like customer service, the selected templates cover $>$70\% of past customer inquiries. Furthermore, the relevance of the model-selected templates is significantly higher than templates selected by a standard tf-idf baseline.

Viaarxiv icon