Catalog - An educational content tagging system.

We present Catalog, an educational content classification and alignment system that tags learning and assessment content in a semantically meaningful and accurate manner. Unlike other approaches that rely on keywords or search terms and crosswalks between knowledge taxonomies, Catalog utilizes powerful NLP, specifically language models based on the Transformer architecture, to encode content in a context attentive fashion. This allows us to capture deep conceptual and contextual relations in content to classify it against a wide variety of educational standards and taxonomies. We present results from empirical studies demonstrating efficacy of our approach in classifying learning content to the Next Generation Science Standards (NGSS).