The Research Of Fuzzy Query Based On Keywords

The expansion of information happened because of the rapid development of information industry, therefore, how to find valid information users needed from the vast amounts of data, becomes a hot research issue. With the increase of data, it becomes increasingly difficult to make an accurate and effective query. Not only some similarities between the keywords exist, but also some wrong inputs, even the 'dirty data' in the database, will bring big difficulties on your query. Therefore, the fuzzy query came into being. At present, many scholars had raised lots of algorithms and thinking about fuzzy query, but most of these algorithms are based on keywords which are exactly matched, and the results they return are often unsatisfactory.In this paper, we presents a new method to solve fuzzy query based on n-gram (the string fragments), which can ensure that the users can get useful information even though they input the wrong queries or there are 'dirty data' in the database. In our paper, we introduce the algorithm and the main thinking of fuzzy query based on the string segmentations. Including the Implementation of string segmentation algorithm, the establishment of inverted index based on B+ tree structure,input pretreatment, finding the candidates from the database using merging algorithm, and finding out the final results from the candidates through edit distance algorithm to the users. And finally, we test and verify the superiority of this idea through experimental platform.