Fusion of Multiple Features for Chinese Organization Names Recognition Based on SVM

In this paper, a hybrid pattern for Chinese organization names based on Support Vector Machine(SVM) is proposed, which fuses multiple features. With consideration of the features of Chinese organization names, local features and global features are abstracted, and feature-vectors are expressed in binary, the training collection is established. From the experimental results on testing set for 1998 peoples´ daily corpus, it can be concluded that the established hybrid model is effective on recognition for Chinese Organization Names. And the experiments on another different testing set also confirm the above conclusion, which shows that this algorithm has consistence on different testing data sources.