Most LLM extraction systems treat each document in isolation, failing to retain and reuse knowledge acquired from previous extractions. That is a problem in domains where terminology changes, taxonomies evolve, and rare concepts matter.
At ACL 2026 in San Diego, we introduce DySECT: a Dynamic Self-Evolving Extraction and Curation Toolkit.
Two Ideas Behind DySECT
Two ideas drive the work:
First, extraction should create reusable knowledge. DySECT uses an LLM to extract structured triples, stores them in a growing knowledge base, and assigns confidence based on source reliability, repeated evidence, and conflicts.
Second, the accumulated knowledge should improve subsequent extraction rounds. The KB enriches previous extractions through hierarchical concept abstractions, confidence calibration, mutual exclusion constraints, and automated property validation. This enriched knowledge is then fed back into the extractor via prompt augmentation, encouraging broader coverage and discovery of new information in later iterations.
Results on DocRED
On DocRED, KB-guided extraction improved recall across GPT-4.1, GPT-4.1-mini, LLaMA-3.3 70B, and Kimi K2.5. The first KB-guided iteration improved recall by 5-8% without fine-tuning or synthetic data generation.
What This Means for Enterprise AI
The broader implication: adaptive extraction does not have to mean opaque model updates. DySECT keeps accumulated knowledge explicit, inspectable, and editable.
For enterprise AI systems, that matters. Knowledge can evolve, but oversight stays in the loop.
Read the research: https://arxiv.org/abs/2603.06915
