Mac Windows Linux

gold-digger

Gold Digger is a command-line tool for snowballing literature search, starting from a few initial papers. It uses the Semantic Scholar API to find citations and references, and provides a flexible filtering mechanism to narrow down the results.

Mac Windows Linux

General and Interdisciplinary

Project Research

Peer Review Research Literature Search CLI

Snowballing Literature Search: Automatically expands a literature search by tracing citations and references from an initial set of papers.

Semantic Scholar API Integration: Fetches academic paper data, including citations and references.

Flexible Filtering System: Allows users to narrow down results using a variety of criteria.

Keyword & Field Filter: Search for keywords within specific paper fields like title or abstract.

Year Filter: Filter papers by publication year using operators (e.g., greater than, equal to).

Author Filter: Filter papers based on an author's name.

LLM-Powered Filter: Utilize a Large Language Model (e.g., Gemini) to filter papers based on a custom, natural language criterion.

File-Based LLM Filter: Provide the LLM filtering criterion via a text file.

Complex Logic (OR): Group filters using OR logic for more flexible searching.

Multiple Output Formats: Save results in JSON or convert them to BibTeX format.

BibTeX Conversion Utility: Convert existing JSON result files to .bib format.

API Response Caching: Caches API calls to improve performance and reduce redundant requests.

Configurable Batch Processing: Control batch sizes for API calls and LLM processing.

Multiple LLM Backends: Supports both the Gemini API and a local gemini-cli tool for LLM filtering.