Determining the character replacement rules and implementing them for phonetic identification of given words to identify similar pronunciation words

摘要

Text pattern recognition is one of the oldest text processing activities in the world of computing. It is easy and the basic word operation to check whether two words are lexically equal or not. Many solutions may exist to determine whether words are equal or not. These solutions are the part of many programming languages in form of procedure or function. To identify two words based on how they are pronounced require more efforts rather than simply testing character by character comparison. The solution to identify phonetic equality between two or more words is somehow tricky. Solution for this problem is achieved by applying a set of rules on the given word. English language is versatile language having many such words also known as homophone describes the rules for phonetic identification. Many such algorithms have been developed with different set of rules they apply. The major objective of the proposed paper is to simplify the process of applicability of such rules. The solution is derived in form of a simplified flow chart which will decide the set of applicable rules. The flow chart process is tested against sample data to determine the correctness.