The Vector Space Model is used in many fields of Information retrieval for its simplification and adaptability. The classical TF-IDF method gets a poor result in Chinese Web information retrieval because it can not describe the documents properly.The N-Level vector model partitions a document into N level text paragraphs. The term weight vectors are defined according to the text paragraphs' context. The N-Level vector model can distinguish the documents more efficiently than the classical TF-IDF method. A Chinese Web information retrieval based on N-Level vector model with modified algorithm will get more significant improvement results.We propose the Chinese Web information retrieval based on N-Level vector model. We designed and realized a Chinese Web information retrieval system. We propose and realize some algorithm on this system such as information Gain, BTWS, query expansion. The results of all algorithms are compared at last. |