Sift-Kg (github.com)

Sift-KG has launched a new tool that allows users to effortlessly convert any collection of documents into a browsable knowledge graph without the need for coding or complex infrastructure. By employing a simple command-line interface (CLI), users can define extraction parameters in YAML format or utilize built-in defaults. The tool automates the processes of entity extraction, duplicate resolution pending user approval, and narrative generation, all of which trace connections across the document set. With features like interactive graph visualization, multi-language support, and support for various LLM providers, Sift-KG caters to a wide array of applications, from investigative journalism to academic research. This tool is significant for the AI/ML community as it democratizes knowledge graph creation, making it accessible to non-technical users while ensuring accuracy through a human-in-the-loop approach. Notably, Sift-KG allows complete control over the merging of entities, linking every piece of extracted information back to its source document for verifiable provenance. It can be used for various domains, including open-source intelligence and legal review, enhancing the reliability of data extraction and analysis. The platform's ability to output versatile formats, such as GraphML and CSV, further supports comprehensive data analysis and integration into existing workflows.