SYSTEMS AND METHODS FOR ANALYZING HEALTH DATA

Systems and methods for efficient information retrieval for clinical data are provided. A document including one or more patient health records can be obtained and tokenized to generate a sequence of tokens, each corresponding to a word or sub-word within the document. A plurality of segments each containing a token sequence comprising a sub-set of the tokens are then generated. Next, a transformer model is used to process each of the plurality of segments to generate a corresponding word-level encoding for each segment. The word-level encodings for each of the segments are combined and fused to obtain document-level contextual data. The combined and fused word-level encodings can then be analyzed, such as to identify named entities and relationships between them.