Skip to main content


Springer Nature is making SARS-CoV-2 and COVID-19 research free. View research | View latest news | Sign up for updates

Table 9 Performance comparisons (F-score) with NN-based systems on the overall-2013 dataset for DDI classification if systems don’t use main processing techniques

From: An attention-based effective neural model for drug-drug interactions extraction

Method Team P(%) R(%) F(%)
NN-based joint AB-LSTM [32] 71.3 66.9 69.3
MCCNN [21] 67.8
Liu CNN [22] 75.3 60.4 67.0
Zhao SCNN [23] 68.5 61.0 64.5
Our model Att-BLSTM 75.9 68.7 71.5
  1. The listed results come from the corresponding papers. The symbol “-” denotes no corresponding values, because the related paper did not provide complete results. Our model only replaces the candidate drugs (row (1) in Table 6), while other systems use basic text processing and replaced candidate drugs (negative instances aren’t filtered). The highest scores are highlighted in bold