Multilingual Dialogue Generation with Shared-Private Memory
This work addresses the lack of cross-lingual feature exploration in dialogue systems, offering a method to enhance multilingual performance, though it is incremental as it builds on existing sequence-to-sequence frameworks.
The paper tackles the problem of monolingual dialogue systems by introducing a multilingual approach using shared-private memory to learn common and language-specific features, resulting in improved performance on Chinese and English corpora, especially with limited training data.
Existing dialog systems are all monolingual, where features shared among different languages are rarely explored. In this paper, we introduce a novel multilingual dialogue system. Specifically, we augment the sequence to sequence framework with improved shared-private memory. The shared memory learns common features among different languages and facilitates a cross-lingual transfer to boost dialogue systems, while the private memory is owned by each separate language to capture its unique feature. Experiments conducted on Chinese and English conversation corpora of different scales show that our proposed architecture outperforms the individually learned model with the help of the other language, where the improvement is particularly distinct when the training data is limited.