The AllMiner project aims to address the current fragmentation in Information Extraction (IE) by proposing a unified conceptual and methodological framework. It highlights the lack of common standards and interoperable tools for structuring knowledge extracted from text.

Its goal is to develop a universal approach based on large language models (LLMs), capable of performing various IE tasks in zero- or few-shot settings without heavy task-specific adaptation. This involves creating a formal ontology of IE, an aligned software library, and a generalizable extraction methodology.

Ultimately, AllMiner will deliver an open and user-friendly platform, making IE accessible to non-specialists while ensuring interoperability, reproducibility, and high-quality results.