Bag-of-Vectors Autoencoders for Unsupervised Conditional Text Generation
This work addresses a bottleneck in text generation for NLP researchers by enabling more effective unsupervised attribute manipulation, though it is incremental as it builds directly on prior methods.
The paper tackles the limitation of single-vector autoencoders in unsupervised conditional text generation by extending the Emb2Emb method to Bag-of-Vectors Autoencoders, which encode text into variable-size bags of vectors to handle longer texts and improve performance, achieving substantially better results than standard autoencoders in unsupervised sentiment transfer.
Text autoencoders are often used for unsupervised conditional text generation by applying mappings in the latent space to change attributes to the desired values. Recently, Mai et al. (2020) proposed Emb2Emb, a method to learn these mappings in the embedding space of an autoencoder. However, their method is restricted to autoencoders with a single-vector embedding, which limits how much information can be retained. We address this issue by extending their method to Bag-of-Vectors Autoencoders (BoV-AEs), which encode the text into a variable-size bag of vectors that grows with the size of the text, as in attention-based models. This allows to encode and reconstruct much longer texts than standard autoencoders. Analogous to conventional autoencoders, we propose regularization techniques that facilitate learning meaningful operations in the latent space. Finally, we adapt Emb2Emb for a training scheme that learns to map an input bag to an output bag, including a novel loss function and neural architecture. Our empirical evaluations on unsupervised sentiment transfer show that our method performs substantially better than a standard autoencoder.