Alternative Data > Best AI Platform
Replicating Human Analysis: Alexandria Text Analytics;
Alexandria Technology specializes in developing classification systems that analyze diverse financial text for entities, sentiment and topics. The company’s algorithms learn contextual understanding by modelling multi-dimensional dependencies across unstructured data, matching financial professional's consensus with ~92% out of sample accuracy (no backfilling, no 'version control').
Built by classification expert Dr. Ruey-Lung Hsiao, Alexandria's classification technology is light years ahead of competition, consistently outperforming bag-of-words NLP with superior out of sample accuracy and superior topic recall.
Alexandria's classifiers are trained by observing financial text manually labelled by buy and sell side financial professionals. From this the programs are able to understand sentiment and topics for company, macro/economic and commodity news, in multiple languages, most recently adding dedicated classifiers for earnings call transcripts and ESG topics.