| Adding Linguistic Knowledge to the Monte-Carlo Framework | ) layer: f(87 a7 d7 yi: f <—Output layer Hidden layer encoding |
| Adding Linguistic Knowledge to the Monte-Carlo Framework | sentence relevance _, Hidden layer encoding / . |
| Adding Linguistic Knowledge to the Monte-Carlo Framework | The first hidden layer contains two disjoint sets of units g’ and 2’ corresponding to linguistic analyzes of the strategy document. |
| Introduction | We employ a multilayer neural network where the hidden layers represent sentence relevance and predicate parsing decisions. |