Using Syntactic Analysis to Increase Efficiency in Visualizing Text Collections

摘要

Self-Organizing Maps (SOMs) are a good method to cluster and visualize large collections of documents, but they are computationally expensive. In this paper, we investigate linguistically motivated reductions on the usual bag-of-words representation, to improve efficiency. We find that reducing the document representation to heads of verb and noun phrases reduces the heavy computational cost without degrading the quality of the map, especially in combination with term reduction techniques. More severe reductions which focus on subject and object nominal phrases are not advantageous.