A Hierarchical Latent Structure for Variational Conversation Modeling

By Yookoon Park, Jaemin Cho, Gunhee Kim. 2018

98 citations recorded by the source corpus.

Variational autoencoders (VAE) combined with hierarchical RNNs have emerged as a powerful framework for conversation modeling. However, they suffer from the notorious degeneration problem, where the decoders learn to ignore latent variables and reduce to vanilla RNNs. We empirically show that this degeneracy occurs mostly due to two reasons. First, the expressive power of hierarchical RNN decoders is often high enough to model the data using only its decoding distributions without relying on the latent variables. Second, the conditional VAE structure whose generation process is conditioned on a context, makes the range of training targets very sparse; that is, the RNN decoders can easily overfit to the training data ignoring the latent variables. To solve the degeneration problem, we propose a novel model named Variational Hierarchical Conversation RNNs (VHCR), involving two key ideas of (1) using a hierarchical structure of latent variables, and (2) exploiting an utterance drop regularization. With evaluations on two datasets of Cornell Movie Dialog and Ubuntu Dialog Corpus, we show that our VHCR successfully utilizes latent variables and outperforms state-of-the-art models for conversation generation. Moreover, it can perform several new utterance control tasks, thanks to its hierarchical latent structure.

Subfields: Artificial Intelligence, Natural Language Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Natural Language Processing

Citations in the corpus

Human-centric Dialog Training via Offline Reinforcement Learning (2020, methodology): “For model architectures, we focused on hierarchical sequence-tosequence models (Serban et al., 2016, 2017b; Park et al., 2018) because they were found to be more effective for the datasets under consideration than e.”
Approximating Interactive Human Evaluation with Self-Play for Open-Domain Dialog Systems (2019, methodology): “We ran a 3-factor ANOVA on the sum of user scores, where the independent variables are model architecture (HRED, VHRED, VHCR), EI regularization (Baseline, EI), and dataset (C ORNELL , R EDDIT ).”
Lagging Inference Networks and Posterior Collapse in Variational Autoencoders (2019, methodology): “…et al., 2018), deploying less powerful de-coders (Yang et al., 2017; Semeniuta et al., 2017), using lossy input (Chen et al., 2017), utilizing different latent variable connections (Dieng et al., 2017; 2018; Park et al., 2018), or changing the prior (Tomczak & Welling, 2018; Xu & Durrett, 2018).”
Mixture Content Selection for Diverse Sequence Generation (2019, background): “…such as language modeling (Bowman et al., 2016), machine translation (Zhang et al., 2016; Su et al., 2018; Deng et al., 2018; Shankar and Sarawagi, 2019), and conversation modeling (Serban et al., 2017; Wen et al., 2017; Zhao et al., 2017; Park et al., 2018; Wen and Luong, 2018; Gu et al., 2019).”
Treating Dialogue Quality Evaluation as an Anomaly Detection Problem (2020, background): “Finally, Park et al. (2018) report that VHRED sufferers from a degeneration of the latent variable, which renders the model to behave almost like an HRED.”
DialogWAE: Multimodal Response Generation with Conditional Wasserstein Auto-Encoder (2018, background): “However, previous studies have shown that VAE models tend to suffer from the posterior collapse problem, where the decoder learns to ignore the latent variable and degrades to a vanilla RNN (Shen et al. , 2018; Park et al. , 2018; Bowman et al. , 2015).”
Improving Dialog Evaluation with a Multi-reference Adversarial Dataset and Large Scale Pretraining (2020, background): “It also correlates well with human judgments on responses generated by ﬁve dialog generation systems (Serban et al., 2016, 2017; Park et al., 2018; Zhang et al., 2020b).”
FlowQA: Grasping Flow in History for Conversational Machine Comprehension (2018, methodology): “Prior work also modeled hierarchical representation of the conversation history (Park et al., 2018; Suhr & Artzi, 2018).”
CONVERSATIONAL MACHINE COMPREHENSION (2019, methodology): “Prior work also modeled hierarchical representation of the conversation history (Park et al., 2018; Suhr & Artzi, 2018).”
Optimus: Organizing Sentences via Pre-trained Modeling of a Latent Space (2020, methodology): “( vi ) VHCR: a hierarchical VAE model with conversation modeling (Park et al., 2018).”