Abstract: Improved systems and methods for named entity recognition (NER) are disclosed and can include attaching domain-specific context to extracted data. In a particular example implementation, the techniques can include an artificial intelligence (AI) based entity extraction and labeling process using unstructured data as input. The generated labels can include automatically determined entity types. The techniques can further include a domain-aware entity resolution process. First, applying a reverse question-and-answer (Q&A) technique to the output of the entity extraction and labeling process can generate a set of predicted entity keys (e.g., predicted metadata identifiers, such as likely database column references) for the extracted entities and entity types. Second, entity alignment operations can enable determining domain-specific entity keys for the predicted entity keys.

Abstract: Systems and methods are disclosed comprising techniques for signal conversion, such as monitoring digital communications signal data transmitted among a user set, converting signal data of a first discrete signal chunk into a first alphanumeric signal set, inputting the first alphanumeric signal set into a machine learning model to generate a first scoring distribution that maps to a first signal event category set, retrieving, for a second discrete signal chunk, a second alphanumeric signal set and a second scoring distribution that maps to a second signal event category set, determining a signal event divergence score via comparing the first and the second scoring distributions, and when the signal event divergence score fails to satisfy a tolerance threshold, transmitting, to at least one user of the user set, an interventive action that, when executed by the at least one user, causes a transformation of the digital communications signal data.

Abstract: Improved systems and methods for named entity recognition (NER) are disclosed and can include attaching domain-specific context to extracted data. In a particular example implementation, the techniques can include an artificial intelligence (AI) based entity extraction and labeling process using unstructured data as input. The generated labels can include automatically determined entity types. The techniques can further include a domain-aware entity resolution process. First, applying a reverse question-and-answer (Q&A) technique to the output of the entity extraction and labeling process can generate a set of predicted entity keys (e.g., predicted metadata identifiers, such as likely database column references) for the extracted entities and entity types. Second, entity alignment operations can enable determining domain-specific entity keys for the predicted entity keys.