Personalized dialogue generation is the task of generating coherent, fluent dialogue consistent with a specific personality. However, in the current approaches, multiple dialogue examples corresponding to the dialogue agent’s persona are still needed to finetune the model.
A recent paper on arXiv.org proposes a novel Dual Latent Variable Generator for personalized dialogue generation. It does not rely on any personality information or corresponding dialogue examples. The response is generated given only the dialogue context.
Contrary to prior frameworks, the proposed one models both the latent distribution over potential dialogue responses as well as the latent distribution over the agent’s potential persona. Also, the researchers introduce a variance regularization technique and lexical diversity selection method to improve the quality of the generated responses in terms of persona consistency and human likeness.
The generation of personalized dialogue is vital to natural and human-like conversation. Typically, personalized dialogue generation models involve conditioning the generated response on the dialogue history and a representation of the persona/personality of the interlocutor. As it is impractical to obtain the persona/personality representations for every interlocutor, recent works have explored the possibility of generating personalized dialogue by finetuning the model with dialogue examples corresponding to a given persona instead. However, in real-world implementations, a sufficient number of corresponding dialogue examples are also rarely available. Hence, in this paper, we propose a Dual Latent Variable Generator (DLVGen) capable of generating personalized dialogue in the absence of any persona/personality information or any corresponding dialogue examples. Unlike prior work, DLVGen models the latent distribution over potential responses as well as the latent distribution over the agent’s potential persona. During inference, latent variables are sampled from both distributions and fed into the decoder. Empirical results show that DLVGen is capable of generating diverse responses which accurately incorporate the agent’s persona.
Research paper: Lee, J. Y., Aik Lee, K., and Gan, W. S., “DLVGen: A Dual Latent Variable Approach to Personalized Dialogue Generation”, 2021. Link: https://arxiv.org/abs/2111.11363