New Sparse Coverage Framework Enhances Patent Prior-Art Retrieval

Researchers have developed a new framework called Sparse Coverage for patent prior-art retrieval. This unsupervised method maps local span embeddings to a sparse vocabulary of embedding-space centers, which are selected using a coverage-oriented k-center objective. Experiments on the CLEF-IP 2013 dataset indicate that Sparse Coverage achieves comparable or superior document-level recall compared to strong dense patent encoders, while also remaining competitive for passage-level retrieval. AI

IMPACT This new retrieval framework could improve the efficiency and accuracy of searching through complex technical documents like patents.

RANK_REASON The cluster contains an academic paper detailing a new method for information retrieval. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• alphaXiv
• arXiv
• CatalyzeX
• CLEF-IP 2013
• DagsHub
• Gotit.pub
• Hugging Face
• ScienceCast
• Sparse Coverage
• You Zuo

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →