Liwanag, Marianne Colline P. and Padua, Ma. Jean Antoine G.
Further enhancement of match rating approach algorithm applied speech-based dictionary system - Undergraduate Thesis: (BSCS major in Computer Science) - Pamantasan ng Lungsod ng Maynila, 2015.
ABSTRACT: The proponents study the idea of phonetic algorithms or algorithms for matching different words with similar pronunciation to the same code, which allows phonetic similarity based word set comparison and indexing. These algorithms are very useful for searching in lists of people in databases, as well as for using in a spell checker. They are often used in providing users with a handy search by name and surname in databases, lists of people and so on. The motivation is to find an algorithm that can use not only on names but also on English words. There are necessarily complex algorithms with many rules and exception, because English spelling and pronunciation is complicated by past changes in pronunciation and words borrowed from many languages. Among these phonetic algorithms, Match Rating Approach algorithm is the most applicable because it provides its own comparison process. However, Match Rating Approach algorithm has some phases that produce wrong simulations or conversions of the word resulting into homophonic words evaluated as unmatched and vice versa. These phases are removing second consonant of any double consonants present and exact match requirement of the algorithm to evaluate words as homophonic. Also, some rules in digraphs, diphthongs and silent letters were not yet implemented. Thus, the proponents enhanced the algorithm and provided more rules to deliver comprehensive results.
5
5
.
QA76.15.L59.2015
Further enhancement of match rating approach algorithm applied speech-based dictionary system - Undergraduate Thesis: (BSCS major in Computer Science) - Pamantasan ng Lungsod ng Maynila, 2015.
ABSTRACT: The proponents study the idea of phonetic algorithms or algorithms for matching different words with similar pronunciation to the same code, which allows phonetic similarity based word set comparison and indexing. These algorithms are very useful for searching in lists of people in databases, as well as for using in a spell checker. They are often used in providing users with a handy search by name and surname in databases, lists of people and so on. The motivation is to find an algorithm that can use not only on names but also on English words. There are necessarily complex algorithms with many rules and exception, because English spelling and pronunciation is complicated by past changes in pronunciation and words borrowed from many languages. Among these phonetic algorithms, Match Rating Approach algorithm is the most applicable because it provides its own comparison process. However, Match Rating Approach algorithm has some phases that produce wrong simulations or conversions of the word resulting into homophonic words evaluated as unmatched and vice versa. These phases are removing second consonant of any double consonants present and exact match requirement of the algorithm to evaluate words as homophonic. Also, some rules in digraphs, diphthongs and silent letters were not yet implemented. Thus, the proponents enhanced the algorithm and provided more rules to deliver comprehensive results.
5
5
.
QA76.15.L59.2015