Content preserving text generation with attribute controls
This work addresses the need for controlled text generation in NLP applications, but it appears incremental as it builds on existing methods with specific enhancements.
The paper tackles the problem of modifying textual attributes in sentences while preserving content, using a model that combines reconstruction and adversarial losses to generate fluent and attribute-compatible sentences, showing improvements over prior methods in evaluations.
In this work, we address the problem of modifying textual attributes of sentences. Given an input sentence and a set of attribute labels, we attempt to generate sentences that are compatible with the conditioning information. To ensure that the model generates content compatible sentences, we introduce a reconstruction loss which interpolates between auto-encoding and back-translation loss components. We propose an adversarial loss to enforce generated samples to be attribute compatible and realistic. Through quantitative, qualitative and human evaluations we demonstrate that our model is capable of generating fluent sentences that better reflect the conditioning information compared to prior methods. We further demonstrate that the model is capable of simultaneously controlling multiple attributes.