This is a reproduction of the study [1]. It uses a SVM to classify documents with more than one label (Multi-label classification). The dataset used in [1] and this program is the Reuters-21578 with 10,788 news documents classified into 90 topics. The output of the program reproduces the Fig 2 in [1].
| accuracy | precision | recall | F1 |
|---|---|---|---|
| 0.80 | 0.87 | 0.85 | 0.85 |
[1] Godbole, Shantanu, and Sunita Sarawagi. "Discriminative methods for multi-labeled classification." Advances in Knowledge Discovery and Data Mining. Springer Berlin Heidelberg, 2004. 22-30.