SYSTEM FOR AUTOMATED ANALYSIS OF CLINICAL TEXT FOR PHARMACOVIGILANCE

A system including a server including a processor, memory coupled with the processor, a plurality of databases coupled with the processor, and a data interface, the memory having stored thereon processor executable instructions including a data virtualization and query module, an ontology framework module, a search framework module, a natural language processing (NLP) pipeline, and a rules engine, where the data virtualization and query module, when executed by the processor, is operable to enable the system to interface with the plurality of databases and to access data from multiple heterogeneous data sources, wherein the plurality of databases are updated and maintained by separate organizations and the system is operable to access the databases, and wherein the data visualization and query module enables the NLP pipeline to access the plurality of databases, the plurality of databases including a medical record database, where the medical record database contains medical information to be processed by the system, and where the medical information includes textual data and is received from different sources and in different formats, a curated reference and clinical knowledge database, a coding database, a causal and temporal model database, and a curated drug safety and case database, the ontology framework module when executed by the processor creates an initial custom ontology, the server configured to periodically scan the medical record database to detect changes or additions to the medical record database and communicate the changes or additions to the ontology framework to augment the initial custom ontology and thereby create a custom ontology, the search framework module, when executed by the processor, enables search capabilities for the system, the NLP pipeline, in connection with the ontology framework module and data virtualization and query module, when executed by the processor accesses one or more of the medical record databases, curated reference and clinical knowledge databases, and the causal and temporal model databases in performing language processing functions, and processes, by applying the custom ontology, the medical information in the medical record database to identify semantic classes for terms in the textual data by matching the terms in the textual data to attributes associated with the semantic classes, wherein the semantic classes include a medical substance class, a medical condition class, and a time class, the time cla