Neural side effect discovery from user credibility and experience-assessed online health discussions

Table 9 Performance of CNN-NEAT’s Attention versus baselines in Side Effect Extraction in term of Precision

Methods	Ibuprofen	Levothyroxine	Metoformin	Omeprazole	Alprazolam
UMLS Tagging	0.6801	0.6145	0.8378	0.5218	0.614
Neural Extractor [21]	0.6741	0.6259	0.8092	0.4665	0.6161
CNN-NEAT’s Attention	0.7073	0.7119	0.8557	0.504	0.688

The Methods column includes the two baselines, UMLS Tagging and Neural Extractor, and the extracted attention of our proposed NEAT – CNN-NEAT’s Attention. We present the five evaluated drugs, Ibuprofen, Levothyroxine, Metoformin, Omeprazole, Alprazolam and the Precision of extracting their side effects for all three methods.

ISSN: 2041-1480