The Research on Automatic Library Classification Model Based on SUMO Ontology
Hu Zewen · Journal of Intelligence · 2011
This paper proposes an automatic library classification model based SUMO(Suggested Upper Merged Ontology) ontology.The model firstly carries a preprocessing to the collected library information,forming the high-dimensionality word vector space,then uses our designed mapping algorithm based on SUMO ontology to transform word vector space into the low-dimensionality concept vector space,and finally,based on this,comes the training of the automatic library classification model.When users input library information,model can make the same processing to these information and form concept vector space,then use the trained classification model to classify them into the corresponding category.Experiment results show that this model can simplify library classification process,reduce vector space dimensionality,improve classification efficiency,effectively ensure the precision and recall of library classification.