Table 4-2.

Representative syndrome classification approaches.

Category	Example approaches	Application
Manual grouping	Medical experts perform the mapping of laboratory test orders into syndrome categories (Ma et al., 2005).	The BioSense system (Bradley et al., 2005; Sokolow et al., 2005) and Syndromal Surveillance Tally Sheet program in EDs of Santa Clara County, California.
Natural language processing (NLP)	NLP-based approaches classify free-text CCs with simplified grammar containing rules for nouns, adjectives, prepositional phrases, and conjunctions. Critiques of NLP-based methods include lack of semantic markings in chief complaints and the amount of training needed.	As part of RODS, Chapman et al. adapted the MPLUS, a Bayesian network-based NLP system, to classify the free-text chief complaints (Chapman et al., 2005; Wagner et al., 2004a).
Bayesian classifiers	Bayesian classifiers, including naïve Bayesian classifiers, bigram Bayes, and their variations, can classify CCs learned from the training data consisting of labeled CCs.	The CoCo Bayesian classifier from the RODS project (Chapman et al., 2003)
Text string searching	A rule-based method that first uses keyword matching and synonym lists to standardize CCs. Predefined rules are then used to classify CCs or ICD-9 codes into syndrome categories.	EARS (Hutwagner et al., 2003), ESSENCE (CDC, 2003), and the National Bioterrorism Syndromic Surveillance Demonstration Program (Yih et al., 2005)
Vocabulary abstraction	This approach creates a series of intermediate abstractions up to a syndrome category from the individual data (e.g., signs) for syndromes due to an agent of bioterrorism.	The BioStorm system (Crubézy et al., 2005; Buckeridge et al., 2002; Shahar and Musen, 1996)
Ontology-based classification	A rule-based system that can generalize symptoms grouping rules based on UMLS-derived vocabularies and semantics. It provides a flexible architecture for changing or adapting new syndromic categories.	The syndromic mapping component of the BioPortal system (Lu et al., 2008)