作者: Yoelle Maarek , Michael Herscovici , Eitan Farchi , David Carmel , Ronald Fagin
DOI:
关键词: Information retrieval 、 Pruning (decision trees) 、 Lossy compression 、 Ranking (information retrieval) 、 Index (economics) 、 Index compression 、 Data mining 、 Computer science
摘要: An apparatus is provided for performing a method (Fig. 2) pruning an index of corpus text documents, wherein the includes steps ranking (50) postings in and (48) from below given level ranking. The methods invention are lossy, since some document removed full index; however, user cannot differentiate lossy index.