PaperSeek
A document search agent based on a large language model, with natural language input, automatically iteratively searches and exports a list of reviewable papers.
** Description:** PaperSeek is a self-hosted LLM document discovery tool for researchers. Enter research questions, research gaps, or search intentions in natural language, and the system automatically executes four pipeline stages: 1. ** Query generation ** -Convert research questions into search queries for data sources. 2. ** Data source request ** -Query OpenAlex, Crossref, or Web of Science (OpenAlex does not require an API Key). 3. ** Metadata sorting ** -Use metadata and summaries to rank results for relevance. 4. ** Literature results ** -Presents a sorted list of papers that can be exported to CSV, supporting manual review. Support citation seed expansion (discovering relevant research through references), configurable iteration times, and citation map visualization tabs. Historical records are saved locally to SQLite. ** Access and login: ** Access is protected by the lazy cat account and no need to login separately. Fill in the LLM API Key (you can choose the provider and model) in the deployment parameters. OpenAlex is the default data source and can be used without a Key.
