Semantic Self-Organizing Map for Natural Disasters News
スポンサーリンク
概要
- 論文の詳細を見る
In this paper we explore the properties of a set of news published after a natural disaster by using SOM (Self-organizing Maps), mainly the semantic characteristics of the generated map. SOM is used to produce a low-dimensional representation of the input data space by mapping high dimensional vectors into a 2-dimensional grid. Our SOM is composed by fewer units than input vectors ; training stage produces not only a visual representation but also a set of quantization points that can be considered as groups of highly related news. In order to get semantic descriptions about map organization, news context is interpreted as a set of most frequent words that gives an idea of their "meanings". The meanings obtained forms areas that divide the map, showing automatically the semantic relation of the inputs. Firstly, online news are converted to text files and encoded as numerical vectors by applying principal component analysis (PCA). Secondly, a SOM is developed and trained. Finally, semantic descriptions are obtained for each quantization point by interpreting the news contexts.
- 一般社団法人電子情報通信学会の論文
- 2013-02-14
著者
-
Yamashita Katsumi
Graduate School Of Engineering Osaka Prefecture University
-
He Cuiwei
Graduate School of Engineering and Science. Department of Information Engineering. University of the Ryukyus.
-
Villa Rafael
Regional Public Goods. InterAmerican Development Bank
-
Gutierrez Carlos
Graduate School of Engineering and Science. Department of Information Engineering. University of the Ryukyus.
-
Alsharif Mohamad
Graduate School of Engineering and Science. Department of Information Engineering. University of the Ryukyus.
-
Miyagi Hayao
Graduate School of Engineering and Science. Department of Information Engineering. University of the Ryukyus.
-
Tamaki Shiro
Graduate School of Engineering and Science. Department of Information Engineering. University of the Ryukyus.
-
Yamashita Katsumi
Graduate School of Engineering. Osaka Prefecture University.
-
Yamashita Katsumi
Graduate School of Engineenng, Osaka Prefecture University
関連論文
- Signal Recovery in Multiuser MIMO-OFDM Systems with Known Partial Channel State Information
- Multiuser data separation for short message service using ICA (通信方式)
- Multiuser data separation for short message service using ICA (回路とシステム)
- Multiuser data separation for short message service using ICA (信号処理)
- Channel Estimation for Mobile OFDM Systems Using CR Splines
- Fractionally Spaced Bayesian Decision Feedback Equalizer(Regular Section)
- PAPR Reduction for PCC-OFDM Systems Using Neural Phase Rotator
- Cluster Map Based Blind RBF Equalizer(Digital Signal Processing)
- An Equalization Technique for High-Speed-Mobile OFDM Systems in Rayleigh Multipath Channels(Fundamental Theories)
- Bayesian Decision Feedback Equalizer with Receiver Diversity Combining(Digital Signal Processing)
- A Cluster Map Based Blind RBF Decision Feedback Equalizer with Reduced Computational Complexity(Digital Signal Processing)
- Complexity Suppression of Neural Networks for PAPR Reduction of OFDM Signal
- A New Robust Watermarking Algorithm for Digital Image Based on DCT-SVD and Arnold Scrambling
- Semantic Self-Organizing Map for Natural Disasters News