Title: Normalised LCS-based method for indexing multidimensional data cube

Authors: Mayank Sharma; Navin Rajpal; B.V. Ramana Reddy; Ravindra Kumar Purwar

Addresses: USIT Guru Gobind Singh Indraprastha University, Sector – 16C Dwarka, Delhi – 110078, India ' USIT Guru Gobind Singh Indraprastha University, Sector – 16C Dwarka, Delhi – 110078, India ' USIT Guru Gobind Singh Indraprastha University, Sector – 16C Dwarka, Delhi – 110078, India ' USIT Guru Gobind Singh Indraprastha University, Sector – 16C Dwarka, Delhi – 110078, India

Abstract: Query processors fail to retrieve information of interest in the presence of misspelled keywords given by users. The above problem persists because most of currently used indexing system does not have fault-tolerance ability to map the misspelled keywords to the correct records stored at physical level of databases. Therefore, the information retrieval systems need additional support of spell check mechanism with limitations for correction of misspelled keywords before submitting them to query processors. In this paper, a data indexing system is introduced for indexing multidimensional data cube, which maps the keywords to the records stored at physical level in multidimensional data structure and also has normalised longest common subsequence-based string approximation method to find correct keywords against misspelled keywords which comes directly to indexing processes through user queries. It provides more than 90% accurate results in mapping misspelled keywords to the physically stored records. These results are consistent even for large datasets.

Keywords: longest common subsequence; normalised LCS; feature distance; string approximation; multidimensional data cube; information retrieval; misspelled keywords; keyword misspelling; fault tolerance; data indexing; user queries.

DOI: 10.1504/IJIIDS.2013.053550

International Journal of Intelligent Information and Database Systems, 2013 Vol.7 No.2, pp.180 - 204

Available online: 26 Apr 2013 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article