An Augmented Transformer Architecture for Natural Language Generation Tasks
This work addresses performance bottlenecks in natural language generation for applications like translation and summarization, but it is incremental as it builds on existing Transformer architectures.
The paper tackled the problem of improving Transformer models for natural language generation by enhancing positional encoding and incorporating linguistic knowledge like POS tagging, resulting in consistently superior performance over the vanilla Transformer in tasks such as automatic translation and summarization.
The Transformer based neural networks have been showing significant advantages on most evaluations of various natural language processing and other sequence-to-sequence tasks due to its inherent architecture based superiorities. Although the main architecture of the Transformer has been continuously being explored, little attention was paid to the positional encoding module. In this paper, we enhance the sinusoidal positional encoding algorithm by maximizing the variances between encoded consecutive positions to obtain additional promotion. Furthermore, we propose an augmented Transformer architecture encoded with additional linguistic knowledge, such as the Part-of-Speech (POS) tagging, to boost the performance on some natural language generation tasks, e.g., the automatic translation and summarization tasks. Experiments show that the proposed architecture attains constantly superior results compared to the vanilla Transformer.