EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

S-Gens: Structure-Aware Synthetic Data Generation for Enhancing Reasoning-Intensive Dense Retrieval

Zhou Lei, Yanqi Xu, Shengbo Chen

Information · 26.04.2026 · 17 (5) · 413

DOI: 10.3390/info17050413 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Information, online erschienen am 26.04.2026. Verfasst von Zhou Lei und 2 weiteren Autor:innen. Die Arbeit nennt 28 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: the Ministry of Education industry–university cooperative education project, Shanghai University.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Dense retrievers rely heavily on high-quality training triplets, yet existing data construction strategies remain inadequate for reasoning-intensive retrieval tasks involving multi-hop reasoning, entity relation tracing, and implicit evidence composition. Positive samples are often based on shallow semantic relevance and fail to capture explicit reasoning chains, while negative samples are typically sampled from lexical overlap or random candidates and therefore provide limited supervision for learning clear decision boundaries. To address these issues, we propose S-Gens, a structure-aware synthetic data generation framework for enhancing reasoning-intensive dense retrieval. S-Gens uses relation paths in an external knowledge graph to synthesize queries and structurally consistent positive samples, and further constructs semantically similar but structurally inconsistent hard negatives. To improve data reliability, we introduce a Siamese graph neural network-based consistency filtering mechanism. Because S-Gens operates entirely during offline supervision construction, it remains model-agnostic, preserves the original inference architecture, and is complementary to graph-guided retrieval or RAG pipelines that inject structure online. Experiments on five benchmark datasets show that S-Gens consistently improves multiple trainable retrievers, with the largest gains on multi-hop reasoning tasks such as WebQSP and HotpotQA. These results indicate that structure-aware synthetic supervision can effectively improve dense retrieval in reasoning-intensive settings.

Förderung

  • the Ministry of Education industry–university cooperative education project
  • Shanghai University

Zitieren

APA

Lei, Z., Xu, Y., & Chen, S. (2026). S-Gens: Structure-Aware Synthetic Data Generation for Enhancing Reasoning-Intensive Dense Retrieval. Information, 17(5), 413. https://doi.org/10.3390/info17050413

Vancouver

Lei Z, Xu Y, Chen S. S-Gens: Structure-Aware Synthetic Data Generation for Enhancing Reasoning-Intensive Dense Retrieval. Information. 2026;17(5):413. doi:10.3390/info17050413

BibTeX
@article{Lei2026,
  author = {Lei, Zhou and Xu, Yanqi and Chen, Shengbo},
  title = {S-Gens: Structure-Aware Synthetic Data Generation for Enhancing Reasoning-Intensive Dense Retrieval},
  journal = {Information},
  year = {2026},
  volume = {17},
  number = {5},
  pages = {413},
  doi = {10.3390/info17050413}
}
Kostenlos starten →