National Journal of System and Information Technology

Payal Pandya

Abstract

Data retrieval is mainly concerned with exact field value matching. But the relevant value may be little bit different than the exact value. In this case, we may lose important data to process. The effectiveness of a retrieval system strongly depends on the result it retrieves. In such cases, users who have knowledge of regular expression can pose query in that way. But generally users who operate applications are not aware of such tricks. e.g. In Student management system, operator is usually of clerical level. In such situations data retrieval may fail even if data exist in system. In this paper, this point is taken into consideration and an algorithm is developed to process and search structured data in order to get all the relevant information. We deal with real world entity names like Employees, Students, Products, etc., rather than considering its meanings and synonyms.

Keywords

Misspelled Queries, Information Retrieval, Matching Similar Data, Similarity Algorithms

Full Text