▎ 摘　　要

NOVELTY - The method involves normalizing an input text into a graphene text. The graphene text is received. A vowel in the graphene text is identified. A first diacritical mark corresponding to the vowel is fetched from a first database having multiple diacritical marks corresponding to multiple vowels. The vowel is replaced with the first diacritical mark to form a diacritically marked text. The first phonemes corresponding to the diacritically marked text is searched within a first category and a speech from the first phoneme is synthesized (770). USE - Method for pre-processing text for conversion to speech in text to speech converter integrated circuit of pre-processor (claimed) using diacritical marks for speech synthesis. ADVANTAGE - The efficiency of extracting the phoneme sequence from the database is improved. The efficiency of a text-to-speech (TTS) is increased by the use of the diacritical marks in the plain text. The efficiency of the text to speech converter is increased. DETAILED DESCRIPTION - INDEPENDENT CLAIMS are included for the following: (1) a pre-processor for converting a text to speech; (2) a system for pre-processing text for conversion to the speech in the text to the speech converter integrated circuit; and (3) an apparatus for pre-processing text for conversion to the speech in the text to the speech converter integrated circuit. DESCRIPTION OF DRAWING(S) - The drawing shows a flowchart illustrating the process of identifying diacritical mark, processing and categorizing phone sequence for diacritically marked text sequence through categorizer. Step for receiving diacritically marked text (710) Step for finding and extracting diacritically marked text (720) Step for creating category for subject diacritical mark with linguistic rule (750) Step for creating phoneme sequence for subject diacritical mark and word sequence (760) Step for synthesizing speech from first phoneme (770)