A team of specialists from the University of the Philippines Diliman, specifically from the field of engineering, computer science, linguistics, speech communication, and music are working on a program that focuses on the modernization of the Filipino language and practices today. Entitled Development of Inter-disciplinary Signal Processing for Pinoy (ISIP) Program: Filipino Vowels and Emotions (FIVE), this program is headed by Dr. Rowena Cristina L. Guevara of the U.P. College of Engineering along with project leaders Dr. Belen Calingación and Dr. Galileo Zafra from the U.P. College of Arts and Letters and Dr. Ramon Acoymo from the U.P. College of Music.
Three projects comprise the first year of the program. Entitled “Towards the Standardization of Filipino: Focus on Vowels of English Loan Words,” the first project examines English loan words and determines the relationship between Filipino and English vowels. Its corpus consists of English loan words which are words in the Filipino vocabulary borrowed from the English language. There have been difficulties in adapting these loan words to the Filipino language, particularly in spelling and pronunciation, due to the difference in the number of vowel sounds used in Filipino and in English: there are five (5) vowel sounds in Filipino while there are twenty (20) in English [1]. An example is the word survey which when used as a loan word can be spelled and pronounced in Filipino as sarbey, serbey, or surbey. The project therefore aims to standardize the vowels used in loan words or to set a definite conversion of an English vowel into a Filipino vowel. The researchers used voice recordings and two neural networks [2]—one developed by the U.P. Department of Computer Science and the other by the Digital Signal Processing Laboratory of the U.P. Department of Electrical and Electronics Engineering— to study the properties of speech and find a basis for a standard mapping of English vowels to Filipino vowels. The Sentro ng Wikang Filipino and the UP Computer Science Department are collaborating on this project.
The second is a joint project of the UP Digital Signal Processing Laboratory and the UP College of Music; it is entitled “Relationships Between Vowel Migration/Modification in Sung Filipino (Tagalog) Texts and Perceived Intelligibility from the Audience Perspective.” Vocal migration is a technique used by singers when performing especially in wide spaces. This technique provides the singers ease of voice production while maintaining the quality and audibility of their voices. However, intelligibility or the comprehensibility of words to the audience is compromised when this technique is used. The project aims to create a guide for classical singers; it uses a GUI (Graphical User Interface) [3]-based software to identify the point at which the quality of sound is attained with minimum effect on its intelligibility.
The last project under the ISIP program is the “Development of Emotion Classification Algorithms for Call Center Speech Analysis,” a joint research of faculty from the Department of Electrical and Electronics Engineering, the Department of Computer Science, and the Department of Speech Communication and Theater Arts. This project aims to assist call centers in monitoring and improving their services through an emotion recognition system. Both the client’s and the call center agent’s emotions are monitored by the system and are classified into anger, boredom, happiness, satisfaction, and neutral emotions. The emotion recognition system is independent of the words uttered. It uses the prosodic features of speech, in particular, the rhythm, pitch contour, and voice parts of the speech, to determine which emotion is captured from the client and from the call center agent during telephone transactions.
The ISIP FIVE program is supported by the UPD Office of the Chancellor through the Open Grant. It is designed to look into the uses of Information Technology in Filipino language, speech, and music. ISIP FIVE is expected to generate not only scholarly publications but also lead to practical solutions that will aid in the standardization of the Filipino language and contribute to future research and technologies specific to the Filipino language. According to the researchers, “The final target is for researchers in Filipino language, speech, music, signal processing and artificial intelligence to get closer to the limits of human perception in source coding these multimedia signals using text and musical score.”
By MMRParreño
References:
1. Guevara, Rowena Cristina L., et al. Development of Inter-Disciplinary Signal Processing for Pinoys (ISIP) Program: Filipino Vowels & Emotions (FIVE). Open Grant Annual Report submitted to the Office of the Vice-Chancellor for Research and Development, University of the Philippines Diliman, 2008.
2. “Neural Network.” Merriam-Webster OnLine Dictionary. 2008. Merriam-Webster, Incorporated. May 2008. <http://www.merriam-ebster.com/dictionary/neural+network>.
3.“Graphical User Interface (GUI).” Encyclopedia Britannica Online. 2008. Encyclopaedia Britannica, Incorporated. May 2008 <http://www.britannica.com/eb/article-9109589/graphical-user-interface>
___________
[1] The English language has 20 distinct vowel sounds according to the TIMIT Speech Corpus, a speech database of broadband recordings of 630 speakers of 8 major dialects of American English, each reading ten phonetically rich sentences.
[2] Neural network is a computer architecture in which a number of processors are interconnected in a manner suggestive of the connections between neurons in a human brain; the network is able to learn through a process of trial and error.
[3] Graphical User Interface (GUI) a computer program that enables a person to communicate with a computer through the use of symbols, visual metaphors, and pointing devices.
