RSMLP: A light Sampled MLP Structure for Incomplete Utterance Rewrite
This addresses the problem of reconstructing conversational utterances for better comprehension, but it appears incremental as it builds on existing IUR methods with a novel lightweight approach.
The paper tackles the Incomplete Utterance Rewriting task by introducing RSMLP, a lightweight MLP-based method with down-sampling, achieving competitive performance on public datasets and in real-world applications.
The Incomplete Utterance Rewriting (IUR) task has garnered significant attention in recent years. Its goal is to reconstruct conversational utterances to better align with the current context, thereby enhancing comprehension. In this paper, we introduce a novel and versatile lightweight method, Rewritten-Sampled MLP (RSMLP). By employing an MLP based architecture with a carefully designed down-sampling strategy, RSMLP effectively extracts latent semantic information between utterances and makes appropriate edits to restore incomplete utterances. Due to its simple yet efficient structure, our method achieves competitive performance on public IUR datasets and in real-world applications.