No abstract

The World Wide Web is growing ever more rapidly, and there are benefits from rich information. Moreover, demands for retrieving similar strings to an input string from dictionaries have been increasing. The edit distance is necessary to retrieve information from a large amount of data using the similarity between two strings. However, drawback of this method is time consumption because the input string must be compared with all strings in dictionaries. This study proposes a new technique for retrieving similar strings from dictionaries at high speed. The method presented can retrieve all similar strings 14 times faster than unigram methods although the edit distance is 3.