What problem does it solve?
Researchers need an efficient way to discover high-quality, year-specific conference papers across major venues. Conf-Papers automates the discovery, filtering, and ranking process so you can focus on reading, synthesis, and knowledge capture rather than manual sifting.
Core Features & Use Cases
- DBLP-based search: retrieves papers from CVPR/ICCV/ECCV/ICLR/AAAI/NeurIPS/ICML for a given year.
- Semantic Scholar enrichment: fills in abstracts and citation counts to improve ranking, and gathers arXiv/DOI details when available.
- Three-dimensional scoring: relevance (to your keywords), popularity (influential citations), and quality (summary-based novelty/experiments) with a top-N result.
- Two-stage filtering: lightweight keyword filtering first, followed by robust re-scoring and filtering for final candidates.
- Note-ready output: generates top papers with metadata for quick note creation in Obsidian, including a note_filename for wikilinks.
Quick Start
Ask conf-papers to fetch the top 10 papers for 2025 across the configured conferences (e.g., ICLR, NeurIPS, ICML, CVPR) and generate a ready-to-note summary.
