Papers by Claire Gardent
Simplicity Level Estimate (SLE): A Learned Reference-Less Metric for Sentence Simplification (2023.emnlp-main)
Copied to clipboard
| Challenge: | Existing evaluation metrics conflate simplicity with correlated attributes such as fluency or meaning preservation. |
| Approach: | They propose a new learning evaluation metric that focuses on simplicity outperforming most existing metrics in terms of correlation with human judgements. |
| Outcome: | The proposed metric outperforms most existing metrics in terms of correlation with human judgements. |
Modeling Global and Local Node Contexts for Text Generation from Knowledge Graphs (2020.tacl-1)
Copied to clipboard
| Challenge: | Recent graph-to-text models generate text from graph data using global or local aggregation . global node encoding allows explicit communication between two distant nodes, but fails to capture long-range relationships. |
| Approach: | They propose to combine global and local aggregation to learn node representations . they propose to use global and locally encoding to learn contextualized node embeddings based on graph data . |
| Outcome: | The proposed models outperform state-of-the-art models on two graph-to-text datasets by 18.01 and 63.69 points. |
Generating Biographies on Wikipedia: The Impact of Gender Bias on the Retrieval-Based Generation of Women Biographies (2022.acl-long)
Copied to clipboard
| Challenge: | Existing efforts to encourage article creation focus on reducing the gender gap in Wikipedia articles. |
| Approach: | They propose a model that retrieves web evidence and generates biographies section by section . they analyze available web evidence to determine the accuracy of the generated text . |
| Outcome: | The proposed model can generate biographies section by section, including citation information, using retrieval mechanisms and a cache-based pre-trained encoder-decoder. |
Augmenting Transformers with KNN-Based Composite Memory for Dialog (2021.tacl-1)
Copied to clipboard
| Challenge: | Recent work has focused on learning architectures with large memories capable of storing external knowledge. |
| Approach: | They propose a method to augment generative Transformer neural networks with information fetching modules. |
| Outcome: | The proposed approach improves performance in generative dialog modeling . external knowledge is retrieved from Wikipedia, images, and human-written dialog utterances . |
Using Local Knowledge Graph Construction to Scale Seq2Seq Models to Multi-Document Inputs (D19-1)
Copied to clipboard
| Challenge: | Current approaches extract portions of web text as input to Sequence-to-Sequence models . a problem is generating relevant knowledge from noisy and redundant input such as webpages . |
| Approach: | They propose to restructure free text into local knowledge graphs that are linearized into sequences . they propose to encode the graph as a sequence and then linearize it into a structured sequence . |
| Outcome: | The proposed method compresses web search information and reduces redundancy . it can be used for open-domain question answering and multi-document summarization tasks . |
Discourse-Based Sentence Splitting (2021.findings-emnlp)
Copied to clipboard
| Challenge: | Sentence splitting is a key component of sentence simplification and has been shown to help human comprehension. |
| Approach: | They propose to use a discourse connective to generate a sentence that is shorter than the input text. |
| Outcome: | The proposed models outperform end-to-end models in learning the various ways of expressing a discourse relation but generate text that is less grammatical. |
Multilingual AMR-to-Text Generation (2020.emnlp-main)
Copied to clipboard
| Challenge: | Existing work on generating text from structured data into English has focused on bridging the gap between structure and natural language (NL) and semantically underspecified input and fully specified output. |
| Approach: | They propose a multilingual approach that can decode into 21 different languages . they leverage advances in cross-lingual embeddings and pretraining to generate multilingual models . |
| Outcome: | The proposed model surpasses baselines that generate into one language in eighteen languages. |
An Error Analysis Framework for Shallow Surface Realization (2021.tacl-1)
Copied to clipboard
| Challenge: | BLEU and METEOR metrics fail to provide information on which linguistic factors impact performance of natural language generation models. |
| Approach: | They propose a framework for error analysis which permits identifying which features of the input affect the models’ results. |
| Outcome: | The proposed framework improves the performance of 174 system runs submitted to the Multilingual SR shared tasks. |
Controllable Sentence Simplification via Operation Classification (2022.findings-naacl)
Copied to clipboard
| Challenge: | Sentence simplification involves a sentence being transformed into a simpler version of itself while preserving its core meaning. |
| Approach: | They propose a controllable-simplification model that tailors simplifications to four global operations . they propose to use a dataset to train highly accurate classification systems for these operations based on syntactic or discourse structure . |
| Outcome: | The proposed model outperforms both end-to-end and controllable approaches in sentence simplification tasks. |
Generating Complex Question Decompositions in the Face of Distribution Shifts (2025.naacl-long)
Copied to clipboard
| Challenge: | Question decomposition has been found to improve large language models’ (LLMs) performance on complex question answering (QA) however, performance on the task remains dominated by supervised approaches, suggesting room for making LLMs better decomposers. |
| Approach: | They propose to generate synthetic decomposition data with only five annotated examples by extending recent advances in using LLM-as-judge and for reranking in novel ways. |
| Outcome: | The proposed approach generates synthetic decomposition data with only five examples over two benchmark datasets. |
KGConv, a Conversational Corpus Grounded in Wikidata (2024.lrec-main)
Copied to clipboard
| Challenge: | a large corpus of 71k English conversations contains on average 8.6 questions . Unlike open domain and task-oriented dialogues, information seeking conversations are driven by the desire to acquire or evaluate knowledge. |
| Approach: | They propose a large corpus of 71k English conversations where each question-answer pair is grounded in a Wikidata fact. |
| Outcome: | The proposed dataset can be used for knowledge-based, conversational question generation . it can also be used to generate single-turn questions from Wikidata triples, question rewriting, question answering from conversation or knowledge graphs and quiz generation. |
Enhancing AMR-to-Text Generation with Dual Graph Representations (D19-1)
Copied to clipboard
| Challenge: | Abstract Meaning Representation (AMR) is a linguistically-grounded semantic formalism that represents the meaning of a sentence as a directed graph. |
| Approach: | They propose a graph-to-sequence model that encodes different but complementary perspectives of the structural information contained in the graph. |
| Outcome: | The proposed model achieves state-of-the-art results on two AMR datasets. |
Multilingual Verbalisation of Knowledge Graphs (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Most work on Knowledge Graph (KG) verbalisation is monolingual leaving open the question of how to scale KG-to-Text generation to languages with varying amounts of resources. |
| Approach: | They explore how to scale KG-to-Text generation to languages with varying resources . they construct multilingual training data and test data for each language . |
| Outcome: | The proposed approach performs best on all 9 languages, compared with other approaches on low vs high resource languages and on in- v. out-of-domain data. |
Context-Aware Document Simplification (2023.findings-acl)
Copied to clipboard
| Challenge: | Recent work on document simplification has focused on sentence-level inputs but fails to preserve the discourse structure. |
| Approach: | They explore various systems that use document context within the simplification process . they investigate the performance and efficiency tradeoffs of system variants . |
| Outcome: | The proposed approach achieves state-of-the-art even when not relying on plan-guidance. |
LORIA / Lorraine University at Multilingual Surface Realisation 2019 (D19-63)
Copied to clipboard
| Challenge: | WO component is easily transferrable between languages, but needs attention for each language separately. |
| Approach: | They present a LORIA / Lorraine University submission to the shared task . they evaluate it on 11 languages covered by the shared challenge . |
| Outcome: | The proposed approach is evaluated on 11 languages covered by the shared task . the main takeaways are that the WO component is easily transferrable between languages . |
Semantic Evaluation of Multilingual Data-to-Text Generation via NLI Fine-Tuning: Precision, Recall and F1 scores (2025.findings-acl)
Copied to clipboard
| Challenge: | KG-to-Text models are prone to errors like Additions and Omissions, and few languages are taken into account since both train and test data are not readily available. |
| Approach: | They propose a multilingual evaluation framework that is reference-less . it allows estimating how much a KG-to-Text Model under- (omission) or over- (addition) generates. |
| Outcome: | The proposed evaluation framework outperforms prior reference-less metrics in correlation with human judgments and provides scores for precision and recall. |
Deep Learning Approaches to Text Production (N18-6)
Copied to clipboard
| Challenge: | Text production is a key component of many NLP applications . Claire Gardent is based in France and is pursuing research in text production . |
| Approach: | This tutorial will cover the fundamentals and state-of-the-art research on neural models for text production. |
| Outcome: | This tutorial will cover the fundamentals and the state-of-the-art research on neural models for text production. |
MuCAL: Contrastive Alignment for Preference-Driven KG-to-Text Generation (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods for KG-to-text generation are limited by the availability of reliable preference data. |
| Approach: | They propose to use a multilingual KG/Text alignment model to generate preference data using three LLMs by ranking candidates and applying Direct Preference Optimization (DPO) on these preferences. |
| Outcome: | The proposed model achieves robust cross-modal retrieval across multiple languages and difficulty levels. |
Generating Questions from Wikidata Triples (2022.lrec-1)
Copied to clipboard
| Challenge: | Existing methods for question generation from knowledge bases rely on extensive pre- and post-processing of the input triple. |
| Approach: | They revisit KBQG using pre training, a new (triple, question) dataset and taking question type into account and provide a more extended KBqg dataset. |
| Outcome: | The proposed approach outperforms existing methods in a standard and in 'zero-shot' setting. |
Document-Level Planning for Text Simplification (2023.eacl-main)
Copied to clipboard
| Challenge: | Existing work on text simplification is limited to sentence-level inputs . attempts to iteratively apply these approaches fail to preserve discourse structure of document . |
| Approach: | They propose a simplification plan that labels each sentence in the input document while considering both its context and internal structure. |
| Outcome: | The proposed model outperforms baselines on two simplification benchmarks and when used to guide document-level simplification models. |
Entity-Based Semantic Adequacy for Data-to-Text Generation (2021.findings-emnlp)
Copied to clipboard
| Challenge: | Existing pre-trained language models have improved the fluency of text generation systems, but semantic adequacy remains an unsolved issue. |
| Approach: | They propose an automatic evaluation metric to assess to what extent models that verbalise RDF graphs produce text that contains mentions of entities occurring in the input. |
| Outcome: | The proposed metric can be used to assess to what extent generation models verbalise RDF graphs produce text that contains mentions of the entities occurring in the input. |
Surface Realisation Using Full Delexicalisation (D19-1)
Copied to clipboard
| Challenge: | Existing approaches to surface realisation model word ordering, morphological inflection and contraction generation are evaluated on 10 languages covered by the SR'18 shared task. |
| Approach: | They propose a modular approach which models each of these components separately and an analysis of the differences in word ordering performance across languages. |
| Outcome: | The proposed model is compared with existing models on 10 languages covered by the SR'18 shared task. |
Multilingual Generation and Answering of Questions from Texts and Knowledge Graphs (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Existing methods for QG-QA are limited to English, but can be used in other languages. |
| Approach: | They propose to bring multilinguality to multimodal QG-QA by using Brazilian Portuguese and Russian data. |
| Outcome: | The proposed approach outperforms a baseline on English and can handle both languages. |
Learning Health-Bots from Training Data that was Automatically Created using Paraphrase Detection and Expert Knowledge (2020.coling-main)
Copied to clipboard
| Challenge: | Current dialog models require large quantities of training data due to privacy issues . current methods require large amounts of training and are even scarcer in the health domain . |
| Approach: | They propose a method for creating dialog corpora from physician-patient interaction data . they combine expert knowledge provided by physicians with automatic paraphrase extraction techniques . |
| Outcome: | The proposed method outperforms existing models in the health domain . it provides a natural way of guiding learning and improves quality of human-machine interactions . |
Generating Syntactic Paraphrases (D18-1)
Copied to clipboard
| Challenge: | Using data-to-text generation, text-totext generation and text reduction, we show that conditioning text generation on syntactic constraints permits the generation of syntakically distinct paraphrases for the same input. |
| Approach: | They propose to use four different models for automatic generation of syntactic paraphrases to study the automatic generation process. |
| Outcome: | The proposed models can generate syntactic paraphrases for the same input and exploit different types of input to increase the number of distinct paraphrased for a given input. |
Automatic and Human-AI Interactive Text Generation (with a focus on Text Simplification and Revision) (2024.acl-tutorials)
Copied to clipboard
| Challenge: | In this tutorial, we focus on text-to-text generation, a class of natural language generation tasks, that takes a piece of text as input and then generates a revision that is improved according to some specific criteria. |
| Approach: | This tutorial focuses on text-to-text generation, a class of natural language generation tasks that takes a piece of text as input and generates a revision that is improved according to some specific criteria. |
| Outcome: | This tutorial focuses on text-to-text generation, a class of natural language generation tasks, that takes a piece of text as input and generates a revision that is improved according to some specificcriteria. |