A Modular Unsupervised Framework for Attribute Recognition from Unstructured Text

TL;DR

This paper introduces POSID, a modular unsupervised framework that extracts structured attributes from unstructured text, demonstrated on incident reports for human attribute recognition without requiring task-specific training.

Contribution

The paper presents a novel unsupervised, adaptable framework that combines lexical and semantic similarity for attribute extraction from unstructured text.

Findings

Effective attribute extraction without supervised training

Demonstrated on incident reports with the InciText dataset

Applicable across different domains

Abstract

We propose POSID, a modular, lightweight and on-demand framework for extracting structured attribute-based properties from unstructured text without task-specific fine-tuning. While the method is designed to be adaptable across domains, in this work, we evaluate it on human attribute recognition in incident reports. POSID combines lexical and semantic similarity techniques to identify relevant sentences and extract attributes. We demonstrate its effectiveness on a missing person use case using the InciText dataset, achieving effective attribute extraction without supervised training.