Table 5.

Five Way Cross-Validation Sentence Classification Results on Structured Abstracts: Using SVMs

	With no windowed features			With windowed features
	Precision	Recall	-score	Precision	Recall	F-score
1575 abs, 21.2 k sents	Accuracy = 79.33%			Accuracy = 84.04%
Aim	0.90	0.84	0.87	0.97	0.97	0.97
Method	0.80	0.82	0.81	0.84	0.82	0.83
Intervention	0.79	0.72	0.75	0.83	0.78	0.80
Results	0.75	0.80	0.77	0.78	0.83	0.80
Conclusion	0.79	0.74	0.77	0.88	0.87	0.87
1740 abs, 22.9 k sents	Accuracy = 85.14%			Accuracy = 91.12%
Aim	0.90	0.83	0.86	0.96	0.95	0.96
Method	0.86	0.91	0.88	0.93	0.95	0.94
Outcome Measure	0.81	0.80	0.80	0.85	0.81	0.83
Results	0.86	0.88	0.87	0.91	0.92	0.92
Conclusion	0.79	0.72	0.75	0.87	0.87	0.87
2280 abs, 29.8 k sents	Accuracy = 80.64%			Accuracy = 85.20%
Aim	0.89	0.82	0.86	0.96	0.95	0.95
Method	0.82	0.85	0.84	0.85	0.85	0.85
Participants	0.80	0.72	0.76	0.87	0.76	0.81
Results	0.76	0.81	0.79	0.79	0.84	0.82
Conclusion	0.82	0.85	0.84	0.88	0.87	0.87

Sentence classification using SVMs into five classes, for each of the three classification problems. Results report using 15-fold cross validation for a system that uses no windowed features versus a system that uses windowed features.