factor oracle,Factor-Oracleを索引に用いた全文検索アルゴリズム

摘要：

A new almost linear time algorithm for indexed full-text search is presented. The algorithm uses a modified factor oracle as an index database. A factor oracle is an acyclic automaton made from an input text, and it recognizes all of substrings in the text. It can be built in time proportional to the text length, and it is more space economical and easy to implement comparing similar machineries such as suffix-tree or suffix-automaton. On the other hand, it has a problem that it may accept strings not in the text ; thsu, we cannot use factor oracles for indexed searching dierctly. We solve this problem and realize full-text-indexing using a modified factor oracle. The search algorithm runs in time proportional to the length of pattern, and it is again easy to implement.

展开