| Algorithmic details | The algorithm classifies publication's fulltexts using a Bayesian classifier and weighted terms according to an offline training phase. The training has been done using the following taxonomies: arxiv, dcc, acm |
| Parameters | Publication's identifier and fulltext |