Referring Expression Generation Using Entity Profiles
This addresses a limitation in REG for natural language processing by enabling generalization to unseen entities, though it is incremental as it builds on existing methods.
The paper tackles the problem of Referring Expression Generation (REG) systems' inability to handle entities not seen during training by proposing a profile-based deep neural network model, ProfileREG, which outperforms baselines on three splits of the WebNLG dataset in automatic and human evaluations.
Referring Expression Generation (REG) is the task of generating contextually appropriate references to entities. A limitation of existing REG systems is that they rely on entity-specific supervised training, which means that they cannot handle entities not seen during training. In this study, we address this in two ways. First, we propose task setups in which we specifically test a REG system's ability to generalize to entities not seen during training. Second, we propose a profile-based deep neural network model, ProfileREG, which encodes both the local context and an external profile of the entity to generate reference realizations. Our model generates tokens by learning to choose between generating pronouns, generating from a fixed vocabulary, or copying a word from the profile. We evaluate our model on three different splits of the WebNLG dataset, and show that it outperforms competitive baselines in all settings according to automatic and human evaluations.