Bootstrapping Generators from Noisy Data (N18-1)

Copied to clipboard

Challenge: Existing methods for data-to-text generation focus on learning correspondences between structured data and associated texts.
Approach: They aim to bootstrap generators from large scale datasets where data and related texts are loosely aligned.
Outcome: The proposed model improves on a vanilla encoder-decoder which relies on soft attention.

Similar Papers

Neural Data-to-Text Generation via Jointly Learning the Segmentation and Correspondence (2020.acl-main)

Copied to clipboard

Challenge: Recent neural attention models conflate all steps into a single end-to-end system and simplify training process.
Approach: They propose to explicitly segment target text into fragment units and align them with their data correspondences.
Outcome: The proposed model outperforms neural attention models on E2E and WebNLG benchmarks.
Data-to-text Generation with Macro Planning (2021.tacl-1)

Copied to clipboard

Challenge: Recent approaches to data-to-text generation adopt the encoder-decoder architecture . however, these models perform poorly at selecting appropriate content and ordering it coherently .
Approach: They propose a neural model with a macro planning stage followed by a generation stage . they use data from databases of records, simulations of physical systems, accounting spreadsheets .
Outcome: The proposed model outperforms baselines on two data-to-text benchmarks . it uses the encoderdecoder architecture and is compared with existing models .
Select and Attend: Towards Controllable Content Selection in Text Generation (D19-1)

Copied to clipboard

Challenge: Recent neural network models conflate content selection and surface realization into a black-box architecture, resulting in content to be described in text cannot be explicitly controlled.
Approach: They propose to decouple content selection from the decoder to allow finer-grained control over the generation.
Outcome: The proposed model can be trained end-to-end without human annotations and achieves promising results in data-totext and headline generation tasks.
Controlled Hallucinations: Learning to Generate Faithfully from Noisy Data (2020.findings-emnlp)

Copied to clipboard

Challenge: Neural text generation (data- or text-to-text) demonstrates remarkable performance when training data is abundant which for many applications is not the case.
Approach: They propose a technique to treat hallucinations as a controllable aspect of the generated text without dismissing any input and without modifying the model architecture.
Outcome: The proposed technique can be used on a WikiBio dataset and in a human evaluation.
Advances in Pre-Training Distributed Word Representations (L18-1)

Copied to clipboard

Challenge: Pre-trained word representations are a building block of many Natural Language Processing and Machine Learning applications.
Approach: They propose to combine known tricks and a set of publicly available pre-trained word vector representations to train high-quality representations.
Outcome: The proposed models outperform the current state of the art on a number of tasks while maintaining a high training speed to scale to massive amount of data.
WikiTableT: A Large-Scale Data-to-Text Dataset for Generating Wikipedia Article Sections (2021.findings-acl)

Copied to clipboard

Challenge: Existing datasets for data-to-text generation focus on single-sentence generation or long-form generation.
Approach: They create a dataset that pairs Wikipedia sections with tabular data and various metadata.
Outcome: The proposed dataset can generate fluent and high quality texts but struggle with coherence and factuality.
Data-to-text Generation by Splicing Together Nearest Neighbors (2021.emnlp-main)

Copied to clipboard

Challenge: Existing work on data-to-text generation relies on retrieved "neighbors" but instead generates text token-by-token, left-to right.
Approach: They propose to splice together retrieved segments of text from "neighbor" source-target pairs to generate text token-by-token, left-to-right.
Outcome: The proposed method performs on par with strong baselines in terms of automatic and human evaluation, but allows for more interpretable and controllable generation.
Little Giants: Synthesizing High-Quality Embedding Data at Scale (2025.naacl-long)

Copied to clipboard

Challenge: Synthetic data generation is an increasingly popular way of training models without the need for large, manually labeled datasets.
Approach: They propose a framework that aligns open-source small models to efficiently generate large-scale embedding data.
Outcome: The proposed framework outperforms state-of-the-art embedding models by using only 1/10 of the GPT API calls.
Domain-matched Pre-training Tasks for Dense Retrieval (2022.findings-naacl)

Copied to clipboard

Challenge: Existing approaches to improve performance of pre-training tasks are needed.
Approach: They propose to pre-train large bi-encoder models on a recently released set of 65 millionsynthetically generated questions and 200 million post-comment pairs from a preexisting reddit conversation dataset.
Outcome: The proposed model can be pre-trained on a set of 65 millionsynthetically generated questions and 200 million post-comment pairs from a preexisting dataset of Reddit conversations.
STable: Table Generation Framework for Encoder-Decoder Models (2024.eacl-long)

Copied to clipboard

Challenge: Existing approaches to infer text-to-table neural models are limited to raw text, but the proposed framework is capable of unifying a variety of problems involving natural language.
Approach: They propose a framework for text-to-table neural models that utilizes a generalized sequential method that comprehends information from all cells in the table.
Outcome: The proposed framework outperforms previous approaches on several challenging datasets and outperformed existing models by up to 15%.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations