Self-supervised Regularization for Text Classification

Conversation Graph: Data Augmentation, Training, and Evaluation for Non-Deterministic Dialogue Management Milan Gritta, Gerasimos Lampouras, Ignacio Iacobacci Huawei Noah’s Ark Lab, London, United Kingdom milan.gritta, gerasimos.lampouras, ignacio.iacobacci@huawei.com Abstract Task-oriented dialogue systems typically rely on large amounts of high-quality training data or require complex handcrafted rules. However, existing datasets are often limited in size con- sidering the complexity of the dialogues. Additionally, conventional training signal in- ference is not suitable for non-deterministic agent behavior, namely, considering multiple actions as valid in identical dialogue states. We propose the Conversation Graph (ConvGraph), a graph-based representation of dialogues that can be exploited for data augmentation, multi- reference training and evaluation of non- deterministic agents. ConvGraph generates novel dialogue paths to augment data vol- ume and diversity. Intrinsic and extrinsic evaluation across three datasets shows that data augmentation and/or multi-reference training with ConvGraph can improve dialogue success rates by up to 6.4%. 1 Introduction Dialogue systems research focuses on the natural language interaction between a user and an artificial conversational agent. Current trends lean towards end-to-end models (Bordes et al., 2017; Miller et al., 2017) while modular systems tend to be the preferred approach in industrial applica- tions (Bocklisch et al., 2017; Burtsev et al., 2018). At the core of a modular conversational agent is dialogue management (DM), whose function is to exchange information with a user, update the agent’s internal state, and plan its next action according to a policy (Young, 2000). The dialogue manager then collects the user’s response and the process repeats until the conversation ends. More specifically, task-oriented dialogue managers are aiming towards the completion of some specific task or tasks (e.g., booking services or buying products). Machine-learned policies for DM require large amounts of high-quality data to generalize to a variety of conversational scenarios (El Asri et al., 2017; Shah et al., 2018; Budzianowski et al., 2018; Rastogi et al., 2019). However, given the com- plexity of some tasks, the datasets are often lim- ited in size. Reinforcement learning approaches (Henderson et al., 2008; Bordes et al., 2017; Miller et al., 2017; Li et al., 2017; Gordon-Hall et al., 2020) can replace the need for explicit training data by exploiting a custom-designed environ- ment to infer the training signal for the policy. Unfortunately, such custom-designed environ- ments may not be representative of how a user would interact with a conversational agent and their manual development is time-consuming and domain-specific. Furthermore, modern conversa- tional agents exhibit non-deterministic behavior, that is, they are able to take different but equally valid actions in identical dialogue states. Con- ventional agent training and evaluation do not support the non-deterministic nature of conver- sational datasets as only a single fixed target is considered per inference, penalizing valid model predictions. To address these problems, we propose the Conversation Graph (ConvGraph), a DM frame- work for data augmentation, which also enables the training and evaluation of non-deterministic policies. If we consider each dialogue as a se- quence of dialogue states, alternating between the agent and the user, we …

Content truncated to 500 words