Injecting Entity Types into Entity-Guided Text Generation

  • Xiangyu Dong ,
  • Wenhao Yu ,
  • Chenguang Zhu ,
  • Meng Jiang

Empirical Methods in Natural Language Processing (EMNLP) 2021 |

Recent successes in deep generative modeling have led to significant advances in natural language generation (NLG). Incorporating entities into neural generation models has demonstrated great improvements by assisting to infer the summary topic and to generate coherent content. In order to enhance the role of entity in NLG, in this paper, we aim to model the entity type in the decoding phase to generate contextual words accurately. We develop a novel NLG model to produce a target sequence (i.e., a news article) based on a given list of entities. The generation quality depends significantly on whether the input entities are logically connected and expressed in the output. Our model has a multi-step decoder that injects the entity types into the process of entity mention generation. It first predicts the token of being a contextual word or an entity, then if an entity, predicts the entity mention. It effectively embeds the entity’s meaning into hidden states, making the generated words precise. Experiments on two public datasets demonstrate type injection performs better than type embedding concatenation baselines.