Diversifying Reply Suggestions using a Matching-Conditional Variational Autoencoder
This addresses the problem of monotonous reply suggestions for users of instant-messaging systems, representing an incremental improvement with a novel method for a known bottleneck.
The paper tackled the problem of diversifying automated reply suggestions in a commercial instant-messaging system (Skype) by formulating a matching-based model as a generative latent variable model with a Conditional Variational Auto-Encoder (M-CVAE), resulting in a ~30-40% increase in diversity without significant impact on relevance and a 5% gain in click-rate in online production.
We consider the problem of diversifying automated reply suggestions for a commercial instant-messaging (IM) system (Skype). Our conversation model is a standard matching based information retrieval architecture, which consists of two parallel encoders to project messages and replies into a common feature representation. During inference, we select replies from a fixed response set using nearest neighbors in the feature space. To diversify responses, we formulate the model as a generative latent variable model with Conditional Variational Auto-Encoder (M-CVAE). We propose a constrained-sampling approach to make the variational inference in M-CVAE efficient for our production system. In offline experiments, M-CVAE consistently increased diversity by ~30-40% without significant impact on relevance. This translated to a 5% gain in click-rate in our online production system.