It is found that the triple composition and the training direction have the greatest impact, whereas the text encoder and hard-negatives sampling matter little and simple choices win: projecting text into the KG space with a linear head over concatenated subject, predicate, and object embeddings performs best.
Abstract
Biomedical knowledge exists in two complementary but distinct forms: unstructured scientific literature and structured knowledge graphs (KGs). Aligning them is essential for knowledge grounding, evidence retrieval, and KG completion, yet existing methods do not explicitly align free-text evidence with KG triples. We present a unified framework for systematically studying design choices for aligning biomedical text and KGs. With a text encoder and a KG embedding model both frozen, we learn only a lightweight projection between their spaces via a contrastive objective. This enables a fair comparison across six design dimensions: text encoder, KG embedding model, projection head, triple composition, training direction, and hard-negatives sampling. We construct CTD-Align, a corpus of over 22K one-to-one tripledocument pairs linking chemical-gene interactions from the Comparative Toxicogenomics Database to supporting PubMed passages. We evaluate alignment on it in two retrieval settings: document-to-triple and triple-to-document. We find that the triple composition and the training direction (i.e., shared retrieval space) have the greatest impact, whereas the text encoder and hard-negatives sampling matter little. Overall, simple choices win: projecting text into the KG space with a linear head over concatenated subject, predicate, and object embeddings performs best. These findings establish lightweight contrastive alignment as an effective, practical foundation for bridging biomedical text and KGs.
Results demonstrate that InfoFlowEX equips LLMs with robust adaptability, achieving consistent gains over baselines with minimal task-specific customization, highlighting InfoFlowEX for real-world biomedical applications.
Wuyang Lan, Siqi Zhang, Wenzheng Wang et al.· Cell Reports Medicine· 0 citations
A three-stage framework made up of neighborhood-aware retrieval, dual reranking, and score fusion that achieves the state of the art on average across five widely-used benchmarks and remains efficient at inference.
Access to objective, reliable scientific information is crucial in a world of misinformation and disinformation, yet the general public often avoids scientific literature due to its perceived complexity. Modern generative information access models hold the promise of removing some of these barriers by developing appropriate approaches to scientific text simplification. We constructed BioCLEAR, a benchmark for Biomedical Comprehensible Lay Explanations of Academic Research used in the CLEF 2025 SimpleText track, comprising a large set of aligned abstracts and plain-language summaries from Cochrane systematic reviews. First, it provides a new resource for scientific document-level text simplification, including discourse-level aspects that are absent from traditional sentence-level data. Second, we include existing resources created through human sentence-level simplification to run experiments across different datasets. Third, it includes a large set of sentence-level and document-level submissions to the CLEF track, providing a sample of current models and enabling detailed analysis of their effectiveness and remaining issues.
Jan Bakker, L. Ermakova, Jaap Kamps· Annual International ACM SIG...· 0 citations
Initial experiments on heart-failure-focused clinical question answering show that CGX improves evidence retrieval quality and perceived answer reliability over conventional retrieval methods, while reducing total graph construction time by 69.7% under the same input corpus and hardware setting.
Dat Nguyen, Anh N. Le, Binh T. D. Trinh et al.· Journal of Biomedical Inform...· 0 citations
: Biomedical texts present significant challenges for natural language processing (NLP) due to their complex terminology, intricate contextual dependencies, and highly domain-specific semantics. This study investigates the effectiveness of knowledge distillation (KD) for biomedical text classification, aiming to develop lightweight, resource-efficient models that remain competitive with larger architectures. A balanced dataset of 25,000 PubMed records was constructed, equally distributed across five biomedical domains. Two teacher models (BERT and PubMedBERT) and five student models (DistilBERT, BioClinicalBERT, BioBERT, DistilBioBERT, and DistilRoBERTa) were evaluated across ten distinct KD configurations. Each student model was also directly fine-tuned to serve as a controlled baseline. Model performance was assessed using accuracy, precision, recall, and F1-score. The results show that KD can, under suitable teacher–student configurations, enable student models to surpass direct fine-tuning, while other configurations yield only marginal or comparable improvements. Among all configurations, BERT → DistilBERT achieved the highest performance, reaching 89% accuracy. Unexpectedly, the general-purpose BERT teacher outperformed the domain-specific PubMedBERT across multiple student models, suggesting that broader linguistic representations can transfer more effectively across diverse biomedical subdomains. Lower performance in certain KD settings, such as DistilBioBERT and DistilRoBERTa, was attributed to architectural mismatches and limited student capacity. These findings demonstrate that compact models can achieve strong biomedical classification performance through KD under compatible teacher–student pairings, while also highlighting that KD effectiveness varies substantially depending on the specific model combination.
Translating complex biomedical data into patient-friendly narratives is central to modern biomedical informatics. This study presents a comparative analysis of training small language models (SLMs) in specialized biomedical datato-text generation tasks. We explore widely adopted post-training methods including supervised fine-tuning (SFT), direct preference optimization (DPO), odds ratio preference optimization (ORPO), and group relative policy optimization (GRPO) with Qwen-based SLMs on a medicine package leaflets dataset. To assess cross-dataset generalizability, we also curated drug label data from openFDA. We evaluate models using both standard lexical overlap metrics like ROUGE as well as semantic similarity measures. Across our experiments, the results show that (1) the aligned SLMs outperform proprietary models like GPT-5; (2) ORPO outperforms the SFTbaselines; (3) GRPO yields the most robust cross-dataset performance among the alignment methods tested as well as GPT-5.
Xi Yang, Guodong Liu, Chuqin Li et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
