Vamstar uses machine learning to filter relevant pages and tables from tender documents, identifying lot references and items for further processing.. TLDR:

Vamstar develops methods to extract domain knowledge from TED tender XMLs, building multilingual lexicons and classifiers for analysis.

Vamstar uses a rule-based system to parse lot references and item information from tender sentences, extracting structured data like item names, forms, and more