Stick to Facts: Towards Fidelity-oriented Product Description Generation

TL;DR

This work tackles fidelity in automatic product description generation by introducing FP­DG, a model that leverages entity-label guidance via an ELSTM decoder and a keyword memory to ensure attribute-faithful output. The core innovations include the ELSTM cell and a memory-augmented attention mechanism that ties generated text to input attribute words. On a large real-world e-commerce dataset, FP­DG achieves state-of-the-art performance on automatic metrics and markedly higher human fidelity scores, with fidelity gains of about 24.6% over a strong baseline. The approach offers a practical path toward error-resistant product descriptions, reducing unfaithful content in online marketplaces.

Abstract

Different from other text generation tasks, in product description generation, it is of vital importance to generate faithful descriptions that stick to the product attribute information. However, little attention has been paid to this problem. To bridge this gap, we propose a model named Fidelity-oriented Product Description Generator (FPDG). FPDG takes the entity label of each word into account, since the product attribute information is always conveyed by entity words. Specifically, we first propose a Recurrent Neural Network (RNN) decoder based on the Entity-label-guided Long Short-Term Memory (ELSTM) cell, taking both the embedding and the entity label of each word as input. Second, we establish a keyword memory that stores the entity labels as keys and keywords as values, allowing FPDG to attend to keywords by attending to their entity labels. Experiments conducted on a large-scale real-world product description dataset show that our model achieves state-of-the-art performance in terms of both traditional generation metrics and human evaluations. Specifically, FPDG increases the fidelity of the generated descriptions by 25%.

Figures (4)

• Figure 1: Overview of FPDG. Green denotes an entity label and purple denotes a word. We divide our model into two components: (1) The Keyword Encoder stores the word and its entity label in the token memory, and uses Self-Attention Modules (SAMs) to encode words and entity labels; (2) The Entity-based Generator generates product description based on the token memory and SAM encoders.
• Figure 2: The structure of ELSTM, which is a hybrid of three LSTMs.
• Figure 3: An overview of the description generator.
• Figure 4: RQ3: Visualizations of entity-label-attention when generating the word on the left.