A paper submitted to the Computer Science > Computation and Language category on October 22, 2026, presents Association Rule Mining as an unsupervised, annotation-free framework for news framing detection. The authors state that computational framing analysis is currently dominated by supervised classifiers, which require large annotated corpora and predefined frame taxonomies.

They claim these requirements limit the use of such methods for non-English languages, novel events, and low-resource settings.