Table 8. SVM model learned using bag-of-words.

Weight	Word
0.98	cellulitis
0.80	uti
0.79	redness_swelling
0.78	sore_throat
0.77	abscess
0.73	diverticulitis
0.72	abscess
0.70	dysuria
0.66	st
0.65	erythema
0.20	swelling
-0.29	swelling_neg
-0.35	pancreatic
-0.36	eye
-0.36	bleed
-0.37	etoh
-0.37	epistaxis
-0.38	pancreatitis
-0.39	injury
-0.57	mvc

Most positive (indicative of infection) and negative (suggesting no infection) words used by the model built by machine learning using the bag-of-words model on triage notes.