Florian Boudin

Mael Houbre, Florian Boudin, Beatrice Daille

Keyphrase generation is the task consisting in generating a set of words or phrases that highlight the main topics of a document. There are few datasets for keyphrase generation in the biomedical domain and they do not meet the expectations in terms of size for training generative models. In this paper, we introduce kp-biomed, the first large-scale biomedical keyphrase generation dataset with more than 5M documents collected from PubMed abstracts. We train and release several generative models and conduct a series of experiments showing that using large scale datasets improves significantly the performances for present and absent keyphrase generation. The dataset is available under CC-BY-NC v4.0 license at https://huggingface.co/ datasets/taln-ls2n/kpbiomed.

Sunisth Kumar, Xanh Ho, Tim Schopf et al.

Multimodal LLMs are increasingly used to assist scientific peer review, where a core requirement is verifying whether claims in a paper are supported by its evidence. Prior work has shown that models perform substantially better at this task when the evidence is a table than when it is a chart of the same underlying data. This raises the question of whether models fail to extract information from charts, or do they extract it but fail to use it when forming their prediction? We study this question through layer-wise linear probing and attention analysis on three open-weight VLMs over table and chart evidence, representing the same underlying data. We find consistent evidence for the latter. Chart information is encoded in the models' intermediate representations but does not reach the prediction position, a gap that is absent for tables and holds across all conditions tested. Attention analysis further reveals that this disconnect takes two architecturally distinct forms across model families. These findings reframe the table-chart gap as a failure of how encoded visual information is routed at prediction time, rather than a failure of encoding itself.

Léane Jourdan, Florian Boudin, Richard Dufour et al.

Writing a scientific article is a challenging task as it is a highly codified genre. Good writing skills are essential to properly convey ideas and results of research work. Since the majority of scientific articles are currently written in English, this exercise is all the more difficult for non-native English speakers as they additionally have to face language issues. This article aims to provide an overview of text revision in writing assistance in the scientific domain. We will examine the specificities of scientific writing, including the format and conventions commonly used in research articles. Additionally, this overview will explore the various types of writing assistance tools available for text revision. Despite the evolution of the technology behind these tools through the years, from rule-based approaches to deep neural-based ones, challenges still exist (tools' accessibility, limited consideration of the context, inexplicit use of discursive information, etc.)

Maël Houbre, Florian Boudin, Beatrice Daille

Keyphrases which are useful in several NLP and IR applications are either extracted from text or predicted by generative models. Contrarily to keyphrase extraction approaches, keyphrase generation models can predict keyphrases that do not appear in a document's text called `absent keyphrases`. This ability means that keyphrase generation models can associate a document to a notion that is not explicitly mentioned in its text. Intuitively, this suggests that for two documents treating the same subjects, a keyphrase generation model is more likely to be homogeneous in their indexing i.e. predict the same keyphrase for both documents, regardless of those keyphrases appearing in their respective text or not; something a keyphrase extraction model would fail to do. Yet, homogeneity of keyphrase prediction models is not covered by current benchmarks. In this work, we introduce a method to evaluate the homogeneity of keyphrase prediction models and study if absent keyphrase generation capabilities actually help the model to be more homogeneous. To our surprise, we show that keyphrase extraction methods are competitive with generative models, and that the ability to generate absent keyphrases can actually have a negative impact on homogeneity. Our data, code and prompts are available on huggingface and github.

Florian Boudin, Akiko Aizawa

Adapting keyphrase generation models to new domains typically involves few-shot fine-tuning with in-domain labeled data. However, annotating documents with keyphrases is often prohibitively expensive and impractical, requiring expert annotators. This paper presents silk, an unsupervised method designed to address this issue by extracting silver-standard keyphrases from citation contexts to create synthetic labeled data for domain adaptation. Extensive experiments across three distinct domains demonstrate that our method yields high-quality synthetic samples, resulting in significant and consistent improvements in in-domain performance over strong baselines.

Léane Jourdan, Julien Aubert-Béduchaud, Yannis Chupin et al.

Scientific writing is an iterative process that generates rich revision traces, yet publicly available resources typically expose only final or near-final versions of papers. This limits empirical study of revision behaviour and evaluation of large language models (LLMs) for scientific writing. We introduce EarlySciRev, a dataset of early-stage scientific text revisions automatically extracted from arXiv LaTeX source files. Our key observation is that commented-out text in LaTeX often preserves discarded or alternative formulations written by the authors themselves. By aligning commented segments with nearby final text, we extract paragraph-level candidate revision pairs and apply LLM-based filtering to retain genuine revisions. Starting from 1.28M candidate pairs, our pipeline yields 578k validated revision pairs, grounded in authentic early drafting traces. We additionally provide a human-annotated benchmark for revision detection. EarlySciRev complements existing resources focused on late-stage revisions or synthetic rewrites and supports research on scientific writing dynamics, revision modelling, and LLM-assisted editing.

Xanh Ho, Yun-Ang Wu, Sunisth Kumar et al.

With the growing number of submitted scientific papers, there is an increasing demand for systems that can assist reviewers in evaluating research claims. Experimental results are a core component of scientific work, often presented in varying formats such as tables or charts. Understanding how robust current multimodal large language models (multimodal LLMs) are at verifying scientific claims across different evidence formats remains an important and underexplored challenge. In this paper, we design and conduct a series of experiments to assess the ability of multimodal LLMs to verify scientific claims using both tables and charts as evidence. To enable this evaluation, we adapt two existing datasets of scientific papers by incorporating annotations and structures necessary for a multimodal claim verification task. Using this adapted dataset, we evaluate 12 multimodal LLMs and find that current models perform better with table-based evidence while struggling with chart-based evidence. We further conduct human evaluations and observe that humans maintain strong performance across both formats, unlike the models. Our analysis also reveals that smaller multimodal LLMs (under 8B) show weak correlation in performance between table-based and chart-based tasks, indicating limited cross-modal generalization. These findings highlight a critical gap in current models' multimodal reasoning capabilities. We suggest that future multimodal LLMs should place greater emphasis on improving chart understanding to better support scientific claim verification.

Julian Schnitzler, Xanh Ho, Jiahao Huang et al.

Most existing multi-hop datasets are extractive answer datasets, where the answers to the questions can be extracted directly from the provided context. This often leads models to use heuristics or shortcuts instead of performing true multi-hop reasoning. In this paper, we propose a new multi-hop dataset, MoreHopQA, which shifts from extractive to generative answers. Our dataset is created by utilizing three existing multi-hop datasets: HotpotQA, 2WikiMultihopQA, and MuSiQue. Instead of relying solely on factual reasoning, we enhance the existing multi-hop questions by adding another layer of questioning that involves one, two, or all three of the following types of reasoning: commonsense, arithmetic, and symbolic. Our dataset is created through a semi-automated process, resulting in a dataset with 1,118 samples that have undergone human verification. We then use our dataset to evaluate five different large language models: Mistral 7B, Gemma 7B, Llama 3 (8B and 70B), and GPT-4. We also design various cases to analyze the reasoning steps in the question-answering process. Our results show that models perform well on initial multi-hop questions but struggle with our extended questions, indicating that our dataset is more challenging than previous ones. Our analysis of question decomposition reveals that although models can correctly answer questions, only a portion - 38.7% for GPT-4 and 33.4% for Llama3-70B - achieve perfect reasoning, where all corresponding sub-questions are answered correctly. Evaluation code and data are available at https://github.com/Alab-NII/morehopqa

Florian Boudin

Citation recommendation is intended to assist researchers in the process of searching for relevant papers to cite by recommending appropriate citations for a given input text. Existing test collections for this task are noisy and unreliable since they are built automatically from parsed PDF papers. In this paper, we present our ongoing effort at creating a publicly available, manually annotated test collection for citation recommendation. We also conduct a series of experiments to evaluate the effectiveness of content-based baseline models on the test collection, providing results for future work to improve upon. Our test collection and code to replicate experiments are available at https://github.com/boudinfl/acm-cr

Florian Boudin, Ygor Gallina, Akiko Aizawa

Sequence-to-sequence models have lead to significant progress in keyphrase generation, but it remains unknown whether they are reliable enough to be beneficial for document retrieval. This study provides empirical evidence that such models can significantly improve retrieval performance, and introduces a new extrinsic evaluation framework that allows for a better understanding of the limitations of keyphrase generation models. Using this framework, we point out and discuss the difficulties encountered with supplementing documents with -- not present in text -- keyphrases, and generalizing models across domains. Our code is available at https://github.com/boudinfl/ir-using-kg

Ygor Gallina, Florian Boudin, Béatrice Daille

Keyphrase generation is the task of predicting a set of lexical units that conveys the main content of a source text. Existing datasets for keyphrase generation are only readily available for the scholarly domain and include non-expert annotations. In this paper we present KPTimes, a large-scale dataset of news texts paired with editor-curated keyphrases. Exploring the dataset, we show how editors tag documents, and how their annotations differ from those found in existing datasets. We also train and evaluate state-of-the-art neural keyphrase generation models on KPTimes to gain insights on how well they perform on the news domain. The dataset is available online at https://github.com/ygorg/KPTimes .

Xanh Ho, Jiahao Huang, Florian Boudin et al.

Extractive reading comprehension question answering (QA) datasets are typically evaluated using Exact Match (EM) and F1-score, but these metrics often fail to fully capture model performance. With the success of large language models (LLMs), they have been employed in various tasks, including serving as judges (LLM-as-a-judge). In this paper, we reassess the performance of QA models using LLM-as-a-judge across four reading comprehension QA datasets. We examine different families of LLMs and various answer types to evaluate the effectiveness of LLM-as-a-judge in these tasks. Our results show that LLM-as-a-judge is highly correlated with human judgments and can replace traditional EM/F1 metrics. By using LLM-as-a-judge, the correlation with human judgments improves significantly, from 0.22 (EM) and 0.40 (F1-score) to 0.85. These findings confirm that EM and F1 metrics underestimate the true performance of the QA models. While LLM-as-a-judge is not perfect for more difficult answer types (e.g., job), it still outperforms EM/F1, and we observe no bias issues, such as self-preference, when the same model is used for both the QA and judgment tasks.

Leane Jourdan, Florian Boudin, Nicolas Hernandez et al.

Writing a scientific article is a challenging task as it is a highly codified and specific genre, consequently proficiency in written communication is essential for effectively conveying research findings and ideas. In this article, we propose an original textual resource on the revision step of the writing process of scientific articles. This new dataset, called CASIMIR, contains the multiple revised versions of 15,646 scientific articles from OpenReview, along with their peer reviews. Pairs of consecutive versions of an article are aligned at sentence-level while keeping paragraph location information as metadata for supporting future revision studies at the discourse level. Each pair of revised sentences is enriched with automatically extracted edits and associated revision intention. To assess the initial quality on the dataset, we conducted a qualitative study of several state-of-the-art text revision approaches and compared various evaluation metrics. Our experiments led us to question the relevance of the current evaluation methods for the text revision task.

Xanh Ho, Anh Khoa Duong Nguyen, An Tuan Dao et al.

The number of Language Models (LMs) dedicated to processing scientific text is on the rise. Keeping pace with the rapid growth of scientific LMs (SciLMs) has become a daunting task for researchers. To date, no comprehensive surveys on SciLMs have been undertaken, leaving this issue unaddressed. Given the constant stream of new SciLMs, appraising the state-of-the-art and how they compare to each other remain largely unknown. This work fills that gap and provides a comprehensive review of SciLMs, including an extensive analysis of their effectiveness across different domains, tasks and datasets, and a discussion on the challenges that lie ahead.

Léane Jourdan, Nicolas Hernandez, Richard Dufour et al.

Revision is a crucial step in scientific writing, where authors refine their work to improve clarity, structure, and academic quality. Existing approaches to automated writing assistance often focus on sentence-level revisions, which fail to capture the broader context needed for effective modification. In this paper, we explore the impact of shifting from sentence-level to paragraph-level scope for the task of scientific text revision. The paragraph level definition of the task allows for more meaningful changes, and is guided by detailed revision instructions rather than general ones. To support this task, we introduce ParaRev, the first dataset of revised scientific paragraphs with an evaluation subset manually annotated with revision instructions. Our experiments demonstrate that using detailed instructions significantly improves the quality of automated revisions compared to general approaches, no matter the model or the metric considered.

Xanh Ho, Sunisth Kumar, Yun-Ang Wu et al.

Scientific claim verification against tables typically requires predicting whether a claim is supported or refuted given a table. However, we argue that predicting the final label alone is insufficient: it reveals little about the model's reasoning and offers limited interpretability. To address this, we reframe table-text alignment as an explanation task, requiring models to identify the table cells essential for claim verification. We build a new dataset by extending the SciTab benchmark with human-annotated cell-level rationales. Annotators verify the claim label and highlight the minimal set of cells needed to support their decision. After the annotation process, we utilize the collected information and propose a taxonomy for handling ambiguous cases. Our experiments show that (i) incorporating table alignment information improves claim verification performance, and (ii) most LLMs, while often predicting correct labels, fail to recover human-aligned rationales, suggesting that their predictions do not stem from faithful reasoning.

Léane Jourdan, Florian Boudin, Richard Dufour et al.

Evaluating text revision in scientific writing remains a challenge, as traditional metrics such as ROUGE and BERTScore primarily focus on similarity rather than capturing meaningful improvements. In this work, we analyse and identify the limitations of these metrics and explore alternative evaluation methods that better align with human judgments. We first conduct a manual annotation study to assess the quality of different revisions. Then, we investigate reference-free evaluation metrics from related NLP domains. Additionally, we examine LLM-as-a-judge approaches, analysing their ability to assess revisions with and without a gold reference. Our results show that LLMs effectively assess instruction-following but struggle with correctness, while domain-specific metrics provide complementary insights. We find that a hybrid approach combining LLM-as-a-judge evaluation and task-specific metrics offers the most reliable assessment of revision quality.

Juan Junqueras, Florian Boudin, May-Myo Zin et al.

Hate speech (HS) is a critical issue in online discourse, and one promising strategy to counter it is through the use of counter-narratives (CNs). Datasets linking HS with CNs are essential for advancing counterspeech research. However, even flagship resources like CONAN (Chung et al., 2019) annotate only a sparse subset of all possible HS-CN pairs, limiting evaluation. We introduce FC-CONAN (Fully Connected CONAN), the first dataset created by exhaustively considering all combinations of 45 English HS messages and 129 CNs. A two-stage annotation process involving nine annotators and four validators produces four partitions-Diamond, Gold, Silver, and Bronze-that balance reliability and scale. None of the labeled pairs overlap with CONAN, uncovering hundreds of previously unlabelled positives. FC-CONAN enables more faithful evaluation of counterspeech retrieval systems and facilitates detailed error analysis. The dataset is publicly available.

Kon Woo Kim, Rezarta Islamaj, Jin-Dong Kim et al.

This study investigates how existing annotation guidelines can be repurposed to instruct large language model (LLM) annotators for text annotation tasks. Traditional guidelines are written for human annotators who internalize training, while LLMs require explicit, structured instructions. We propose a moderation-oriented guideline repurposing method that transforms guidelines into clear directives for LLMs through an LLM moderation process. Using the NCBI Disease Corpus as a case study, our experiments show that repurposed guidelines can effectively guide LLM annotators, while revealing several practical challenges. The results highlight the potential of this workflow to support scalable and cost-effective refinement of annotation guidelines and automated annotation.

Florian Boudin, Akiko Aizawa

Keyphrase generation refers to the task of producing a set of words or phrases that summarises the content of a document. Continuous efforts have been dedicated to this task over the past few years, spreading across multiple lines of research, such as model architectures, data resources, and use-case scenarios. Yet, the current state of keyphrase generation remains unknown as there has been no attempt to review and analyse previous work. In this paper, we bridge this gap by presenting an analysis of over 50 research papers on keyphrase generation, offering a comprehensive overview of recent progress, limitations, and open challenges. Our findings highlight several critical issues in current evaluation practices, such as the concerning similarity among commonly-used benchmark datasets and inconsistencies in metric calculations leading to overestimated performances. Additionally, we address the limited availability of pre-trained models by releasing a strong PLM-based model for keyphrase generation as an effort to facilitate future research.

Enrico Benedetti, Akiko Aizawa, Florian Boudin

Providing example sentences that are diverse and aligned with learners' proficiency levels is essential for fostering effective language acquisition. This study examines the use of Pre-trained Language Models (PLMs) to produce example sentences targeting L2 Japanese learners. We utilize PLMs in two ways: as quality scoring components in a retrieval system that draws from a newly curated corpus of Japanese sentences, and as direct sentence generators using zero-shot learning. We evaluate the quality of sentences by considering multiple aspects such as difficulty, diversity, and naturalness, with a panel of raters consisting of learners of Japanese, native speakers -- and GPT-4. Our findings suggest that there is inherent disagreement among participants on the ratings of sentence qualities, except for difficulty. Despite that, the retrieval approach was preferred by all evaluators, especially for beginner and advanced target proficiency, while the generative approaches received lower scores on average. Even so, our experiments highlight the potential for using PLMs to enhance the adaptability of sentence suggestion systems and therefore improve the language learning journey.

Florian Boudin, Akiko Aizawa

The rapid growth of scholarly literature makes it increasingly difficult for researchers to keep up with new knowledge. Automated tools are now more essential than ever to help navigate and interpret this vast body of information. Scientific papers pose unique difficulties, with their complex language, specialized terminology, and diverse formats, requiring advanced methods to extract reliable and actionable insights. Large language models (LLMs) offer new opportunities, enabling tasks such as literature reviews, writing assistance, and interactive exploration of research. This special issue of the TAL journal highlights research addressing these challenges and, more broadly, research on natural language processing and information retrieval for scholarly and scientific documents.

Julien Aubert-Béduchaud, Florian Boudin, Béatrice Daille et al.

Familiarizing oneself with a new scientific field and its existing literature can be daunting due to the large amount of available articles. Curated lists of academic references, or reading lists, compiled by experts, offer a structured way to gain a comprehensive overview of a domain or a specific scientific challenge. In this work, we introduce ACL-rlg, the largest open expert-annotated reading list dataset. We also provide multiple baselines for evaluating reading list generation and formally define it as a retrieval task. Our qualitative study highlights the fact that traditional scholarly search engines and indexing methods perform poorly on this task, and GPT-4o, despite showing better results, exhibits signs of potential data contamination.

Mael Houbre, Florian Boudin, Beatrice Daille et al.

State-of-the-art models for keyphrase generation require large amounts of training data to achieve good performance. However, obtaining keyphrase-labeled documents can be challenging and costly. To address this issue, we present a self-compositional data augmentation method. More specifically, we measure the relatedness of training documents based on their shared keyphrases, and combine similar documents to generate synthetic samples. The advantage of our method lies in its ability to create additional training samples that keep domain coherence, without relying on external data or resources. Our results on multiple datasets spanning three different domains, demonstrate that our method consistently improves keyphrase generation. A qualitative analysis of the generated keyphrases for the Computer Science domain confirms this improvement towards their representativity property.

Florian Boudin, Béatrice Daille, Evelyne Jacquey et al.

Scientific digital libraries play a critical role in the development and dissemination of scientific literature. Despite dedicated search engines, retrieving relevant publications from the ever-growing body of scientific literature remains challenging and time-consuming. Indexing scientific articles is indeed a difficult matter, and current models solely rely on a small portion of the articles (title and abstract) and on author-assigned keyphrases when available. This results in a frustratingly limited access to scientific knowledge. The goal of the DELICES project is to address this pitfall by exploiting semantic relations between scientific articles to both improve and enrich indexing. To this end, we will rely on the latest advances in semantic representations to both increase the relevance of keyphrases extracted from the documents, and extend indexing to new terms borrowed from semantically similar documents.

Florian Boudin, Ygor Gallina

Neural keyphrase generation models have recently attracted much interest due to their ability to output absent keyphrases, that is, keyphrases that do not appear in the source text. In this paper, we discuss the usefulness of absent keyphrases from an Information Retrieval (IR) perspective, and show that the commonly drawn distinction between present and absent keyphrases is not made explicit enough. We introduce a finer-grained categorization scheme that sheds more light on the impact of absent keyphrases on scientific document retrieval. Under this scheme, we find that only a fraction (around 20%) of the words that make up keyphrases actually serves as document expansion, but that this small fraction of words is behind much of the gains observed in retrieval effectiveness. We also discuss how the proposed scheme can offer a new angle to evaluate the output of neural keyphrase generation models.

Kenichi Iwatsuki, Florian Boudin, Akiko Aizawa

Formulaic expressions, such as 'in this paper we propose', are helpful for authors of scholarly papers because they convey communicative functions; in the above, it is showing the aim of this paper'. Thus, resources of formulaic expressions, such as a dictionary, that could be looked up easily would be useful. However, forms of formulaic expressions can often vary to a great extent. For example, 'in this paper we propose', 'in this study we propose' and 'in this paper we propose a new method to' are all regarded as formulaic expressions. Such a diversity of spans and forms causes problems in both extraction and evaluation of formulaic expressions. In this paper, we propose a new approach that is robust to variation of spans and forms of formulaic expressions. Our approach regards a sentence as consisting of a formulaic part and non-formulaic part. Then, instead of trying to extract formulaic expressions from a whole corpus, by extracting them from each sentence, different forms can be dealt with at once. Based on this formulation, to avoid the diversity problem, we propose evaluating extraction methods by how much they convey specific communicative functions rather than by comparing extracted expressions to an existing lexicon. We also propose a new extraction method that utilises named entities and dependency structures to remove the non-formulaic part from a sentence. Experimental results show that the proposed extraction method achieved the best performance compared to other existing methods.

Ygor Gallina, Florian Boudin, Béatrice Daille

Keyphrase extraction models are usually evaluated under different, not directly comparable, experimental setups. As a result, it remains unclear how well proposed models actually perform, and how they compare to each other. In this work, we address this issue by presenting a systematic large-scale analysis of state-of-the-art keyphrase extraction models involving multiple benchmark datasets from various sources and domains. Our main results reveal that state-of-the-art models are in fact still challenged by simple baselines on some datasets. We also present new insights about the impact of using author- or reader-assigned keyphrases as a proxy for gold standard, and give recommendations for strong baselines and reliable benchmark datasets.

Florian Boudin

We propose an unsupervised keyphrase extraction model that encodes topical information within a multipartite graph structure. Our model represents keyphrase candidates and topics in a single graph and exploits their mutually reinforcing relationship to improve candidate ranking. We further introduce a novel mechanism to incorporate keyphrase selection preferences into the model. Experiments conducted on three widely used datasets show significant improvements over state-of-the-art graph-based models.

Adrien Bougouin, Florian Boudin, Béatrice Daille

Keyphrase annotation is the task of identifying textual units that represent the main content of a document. Keyphrase annotation is either carried out by extracting the most important phrases from a document, keyphrase extraction, or by assigning entries from a controlled domain-specific vocabulary, keyphrase assignment. Assignment methods are generally more reliable. They provide better-formed keyphrases, as well as keyphrases that do not occur in the document. But they are often silent on the contrary of extraction methods that do not depend on manually built resources. This paper proposes a new method to perform both keyphrase extraction and keyphrase assignment in an integrated and mutual reinforcing manner. Experiments have been carried out on datasets covering different domains of humanities and social sciences. They show statistically significant improvements compared to both keyphrase extraction and keyphrase assignment state-of-the art methods.

Florian Boudin, Hugo Mougard, Damien Cram

The SemEval-2010 benchmark dataset has brought renewed attention to the task of automatic keyphrase extraction. This dataset is made up of scientific articles that were automatically converted from PDF format to plain text and thus require careful preprocessing so that irrevelant spans of text do not negatively affect keyphrase extraction performance. In previous work, a wide range of document preprocessing techniques were described but their impact on the overall performance of keyphrase extraction models is still unexplored. Here, we re-assess the performance of several keyphrase extraction models and measure their robustness against increasingly sophisticated levels of document preprocessing.