A novel approach to solve exact matching problem using multi-splitting of text patterns

Abstract

Matching approaches based on the exact match are popular and commonly utilized in various applications such as texting, intrusion detection network systems, web-based search engines and molecular biology. The efficiency of these approaches is measured in terms of time taken in the text searching and effective utilization of heap memory. This paper suggests a unique approach for achieving both time proficiency and better memory usage by splitting the query text pattern used for searching the text Corpus. The propound approach divides the query pattern P into multiple parts such as ${\text{P}}_{{\text{k}}} ,\;{\text{P}}_{{{\text{k}} - {1}}} ,\; \ldots ,\;{\text{P}}_{{2}}$ and P₁ where k depends on the pattern length. In this article, the scanning of the text corpus is performed from the right to the left and the searching of the multiple sub-patterns is performed in the right to the left order. The propound approach applies the traditional Boyer Moore approach to minimize the comparison cost by using a bad match table for the characters of the first sub-pattern P₁ only. The sub-patterns other than P₁ are matched using the Brute Force approach. The sub-patterns are mapped at the beginning of other sub-patterns such as ${\text{P}}_{{\text{k}}} {\text{P}}_{{{\text{k}} - {1}}} \ldots {\text{P}}_{{2}} {\text{P}}_{{1}}$, to find the exact matching. The comparative study of the traditional approaches and the suggested solution indicates that the suggested approach outruns the traditional approaches in terms of memory utilization and matching time.