Dragomir Radev
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Dragomir Radev has a total of 142 co-authors. Here are the top co-authors by number of shared publications:
- Rui Zhang
Yale University
11 shared publications - Caiming Xiong
♠ University of Hong Kong ♣ Salesforce Research
9 shared publications - Davood Rafiei
University of Alberta Edmonton AB Canada
8 shared publications - Alexander R. Fabbri
Yale University Salesforce Research
7 shared publications - Tao Yu
The University of Hong Kong
7 shared publications - Yunyao Li
Adobe Experience Platform
7 shared publications
Latest Publications
- Open Access
- 10 January 2024
- Published by Springer Nature
- p. C1
Correction to: Natural Language Interfaces to Databases
- 10 January 2024
- Published by Springer Nature
- p. C1
Open AccessCorrection to: Natural Language Interfaces to Databases
- 25 November 2023
- Published by Springer Nature
- p. 1-10
Overview
- 25 November 2023
- Published by Springer Nature
- p. 133-175
Data-to-Text
- 25 November 2023
- Published by Springer Nature
- p. 11-25
Building an NLIDB: The Basics
- 25 November 2023
- Published by Springer Nature
- p. 49-103
Text to Data
- 25 November 2023
- Published by Springer Nature
- p. 177-229
Interactivity
- 25 November 2023
- Published by Springer Nature
- p. 105-131
Evaluation
- 25 November 2023
- Published by Springer Nature
- p. 27-47
Data and Query Model
- Open Access
Towards Interpretable and Efficient Automatic Reference-Based Summarization Evaluation
Interpretability and efficiency are two important considerations for the adoption of neural automatic metrics. In this work, we develop strong-performing automatic metrics for reference-based summarization evaluation, based on a two-stage evaluation pipeline that first extracts basic information units from one text sequence and then checks the extracted units in another sequence. The metrics we developed include two-stage metrics that can provide high interpretability at both the fine-grained unit level and summary level, and one-stage metrics that achieve a balance between efficiency and interpretability. We make the developed tools publicly available at https://github.com/Yale-LILY/AutoACU.
Open AccessTowards Interpretable and Efficient Automatic Reference-Based Summarization Evaluation
Interpretability and efficiency are two important considerations for the adoption of neural automatic metrics. In this work, we develop strong-performing automatic metrics for reference-based summarization evaluation, based on a two-stage evaluation pipeline that first extracts basic information units from one text sequence and then checks the extracted units in another sequence. The metrics we developed include two-stage metrics that can provide high interpretability at both the fine-grained unit level and summary level, and one-stage metrics that achieve a balance between efficiency and interpretability. We make the developed tools publicly available at https://github.com/Yale-LILY/AutoACU.
- Open Access
On Improving Summarization Factual Consistency from Natural Language Feedback
Despite the recent progress in language generation models, their outputs may not always meet user expectations. In this work, we study whether informational feedback in natural language can be leveraged to improve generation quality and user preference alignment. To this end, we consider factual consistency in summarization, the quality that the summary should only contain information supported by the input documents, as the user-expected preference. We collect a high-quality dataset, DeFacto, containing human demonstrations and informational natural language feedback consisting of corrective instructions, edited summaries, and explanations with respect to the factual consistency of the summary. Using our dataset, we study three natural language generation tasks: (1) editing a summary by following the human feedback, (2) generating human feedback for editing the original summary, and (3) revising the initial summary to correct factual errors by generating both the human feedback and edited summary. We show that DeFacto can provide factually consistent human-edited summaries and further insights into summarization factual consistency thanks to its informational natural language feedback. We further demonstrate that fine-tuned language models can leverage our dataset to improve the summary factual consistency, while large language models lack the zero-shot learning ability in our proposed tasks that require controllable text generation.
Open AccessOn Improving Summarization Factual Consistency from Natural Language Feedback
Despite the recent progress in language generation models, their outputs may not always meet user expectations. In this work, we study whether informational feedback in natural language can be leveraged to improve generation quality and user preference alignment. To this end, we consider factual consistency in summarization, the quality that the summary should only contain information supported by the input documents, as the user-expected preference. We collect a high-quality dataset, DeFacto, containing human demonstrations and informational natural language feedback consisting of corrective instructions, edited summaries, and explanations with respect to the factual consistency of the summary. Using our dataset, we study three natural language generation tasks: (1) editing a summary by following the human feedback, (2) generating human feedback for editing the original summary, and (3) revising the initial summary to correct factual errors by generating both the human feedback and edited summary. We show that DeFacto can provide factually consistent human-edited summaries and further insights into summarization factual consistency thanks to its informational natural language feedback. We further demonstrate that fine-tuned language models can leverage our dataset to improve the summary factual consistency, while large language models lack the zero-shot learning ability in our proposed tasks that require controllable text generation.