WEAKLY MONITORED DETECTION OF A SEMANTIC UNIT USING COMMON KNOWLEDGE AND TARGET DOMAIN KNOWLEDGE

摘要

Methods and systems for performing a recognition of a semantic unit. The method comprises accessing a document stored in a memory and selecting, from a directory of general knowledge, target domain information based on a specified target domain. The method also includes generating a plurality of weak annotations for the document based on the selected target domain information and the knowledge from a data dictionary of domain specific knowledge and applying the plurality of weak annotations to the document to create a plurality of weak labels. The method further comprises selecting at least one weak label from the plurality of weak labels as training data and training a predictive model of a semantic unit using the training data.