Classification and retrieval of patient records using natural language: An experimental application of latent semantic analysis

The National Library of Medicine's Unified Medical Language System (UMLS) provides a structured lexicon enabling the application of latent semantic analysis for the classification and retrieval of patient diagnoses. Information matrices of complex number values from UMLS entries are constructed to create principal components via singular value decomposition. Natural language diagnosis entries or inquiries can be projected into the resultant N-dimension concept space, and evaluated by cosine deviation form the compressed concept components. Preliminary evaluations show that the technique is promising. A major advantage is the avoidance of manually constructed semantic network data schemes; semantic properties derive from statistical decomposition.