DocumentCode :
3641228
Title :
Hands-free messaging application (iSay-SMS): A proposed framework
Author :
Lai Mei Lee;Umi Kalsom Yusof
Author_Institution :
School of Computer Sciences, Universiti Sains Malaysia, 11800 USM Penang, Malaysia
fYear :
2010
Firstpage :
1126
Lastpage :
1131
Abstract :
Mobile phones have become a necessity in people´s life today as almost everyone will have at least one with them. Short message service (SMS) which is an important component in mobile phones is growing popular among mobile phone subscribers. The use of SMS nowadays is more than the calling function itself. In order to make the SMS function in mobile phones more efficient and attractive, speech recognition which allows users to use it for voice messages and enables users to speak in a natural way to communicate with others through their mobile phones is introduced. However, speech recognition technology to SMS is in the form of standard language is contradicted with the popularity of SMS that is used in the form of text abbreviation or short form. The research was conducted to find a solution for speech recognition technologies to enable it to transfer human speech into personalized text abbreviations. Studies and findings had been conducted regarding the possible technologies to solve problems such as voice-tag, architecture of ASR (Automatic Speech Recognition) system and Part-of-speech (POS). The problem with the inability to transfer speech to text abbreviation will be solved by adding a personalized data dictionary that enables users to save their own personalized words by integrating the voice-tag and part-of-speech tagging (POS) technology into iSay-SMS to help in saving and retrieving the text abbreviations. In addition, proposed framework of iSay-SMS is designed in such a way that promotes its marketability and increases the acceptance among the mobile phone subscribers.
Keywords :
"Speech recognition","Speech","Mobile handsets","Tagging","Hidden Markov models","Dictionaries","Cellular phones"
Publisher :
ieee
Conference_Titel :
Science and Social Research (CSSR), 2010 International Conference on
Print_ISBN :
978-1-4244-8987-9
Type :
conf
DOI :
10.1109/CSSR.2010.5773701
Filename :
5773701
Link To Document :
بازگشت