论文信息 - Two-Level Neural Network for Multi-label Document Classification

Two-Level Neural Network for Multi-label Document Classification

This paper deals with multi-label document classification using neural networks. We propose a novel neural network which is composed of two sub-nets: the first one estimates the scores for all classes, while the second one determines the number of classes assigned to the document. The proposed approach is evaluated on Czech and English standard corpora. The experimental results show that the proposed method is competitive with state of the art on both languages.

Ladislav Lenc | Pavel Král | Ladislav Lenc | P. Král

[1] Zhi-Hua Zhou,et al. Multilabel Neural Networks with Applications to Functional Genomics and Text Categorization , 2006, IEEE Transactions on Knowledge and Data Engineering.

[2] Ladislav Lenc,et al. Deep Neural Networks for Czech Multi-label Document Classification , 2016, CICLing.

[3] Geoffrey E. Hinton,et al. Rectified Linear Units Improve Restricted Boltzmann Machines , 2010, ICML.

[4] Johannes Fürnkranz,et al. Large-Scale Multi-label Text Classification - Revisiting Neural Networks , 2013, ECML/PKDD.

[5] Yiming Yang,et al. Multilabel classification with meta-level features , 2010, SIGIR.

[6] Fabrizio Sebastiani,et al. Machine learning in automated text categorization , 2001, CSUR.

[7] Quoc V. Le,et al. Distributed Representations of Sentences and Documents , 2014, ICML.

[8] Razvan Pascanu,et al. Theano: A CPU and GPU Math Compiler in Python , 2010, SciPy.

[9] Jun Zhao,et al. Recurrent Convolutional Neural Networks for Text Classification , 2015, AAAI.

[10] Malik Yousef,et al. One-class document classification via Neural Networks , 2007, Neurocomputing.

[11] Jimmy Ba,et al. Adam: A Method for Stochastic Optimization , 2014, ICLR.

[12] Nitish Srivastava,et al. Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[13] Bowen Zhou,et al. Improved Neural Network-based Multi-label Classification with Better Initialization Leveraging Label Co-occurrence , 2016, NAACL.