A large-scale pipeline for LLM-assisted corpus annotation: variation and change in the English consider construction

Name: A large-scale pipeline for LLM-assisted corpus annotation: variation and change in the English consider construction
Uploaded: 2026-10-16T15:23:58.490473+00:00
Duration: 610 s
Description: Researchers developed a four-phase pipeline using large language models (LLMs) to automatically annotate 143,933 'consider' constructions in the Corpus of Historical American English (COHA). The method achieved over 98% accuracy and revealed previously undocumented genre-specific changes in how this grammatical construction has evolved over time.

Researchers developed a four-phase pipeline using large language models (LLMs) to automatically annotate 143,933 'consider' constructions in the Corpus of Historical American English (COHA). The method achieved over 98% accuracy and revealed previously undocumented genre-specific changes in how this grammatical construction has evolved over time.

Open the full episode