NTCIR-4 WEB Experiments at Osaka Kyoiku University - Static/Dynamic Scoring Using Link Structure Analysis and Web Page Grouping -
Hitoshi Nakakubo, Peng Zhang, Takashi G. Sato · NTCIR · 2004
Abstract key word We did gram-based indexing and the retrieval with NTCIR-4 WEB task. The time required to make indices are 34.7 hours.The size of indices is r index technique of WEB contents30.2Gbyte. The median of retrieval time par word is 26msec. The ranking algorithm of retrieval results is based on a traditional probabilistic model. We report on the result of gram-based indexing and the retrieval, and propose a scoring method based on link structure analysis. Keywords: NTCIR, Web Retrieval, Link Structure Analysis, Gram-based Index specifications to experiment. 1. Introduction uNowadays, it is the most general method to use search engines to retrieve aimed WEB pages. WEB pages including key words are expected to be extracted by inputting them to search engines when we retrieve WEB pages. This can be achieved by using full -text retrieval system s. We report on construction of a full-text retrieval system using gram-based indices, and propose scoring method based on link structure analysis in this paper. We describe gram-based indexing and retrieval in sections 2 through 5. And, we propose scoring method based on link structure analysis in sections 6 through 8, and describe conclusion s in the last section. '