I am interested in the unique signals that humans produce and perceive in spoken language, that aren't available in other kinds of language. For example, what does the tone of your voice say about your mood, how sarcastic you're being, whether or not you agree with or like the person your talking to, or if you want to keep talking?
I believe that the best way to answer these questions is to leverage mathematical tendencies of language using computational techniques. This means that we create statistical models of language use and test them on real-world data. Computational techniques also lend themselves to applications in speech technology. For instance, in concert with audiological collaborators I have designed assessments that can be used to tune self-fitting hearing aids.
Computational Speech Lab
The majority of my research is conducted alongiside the hardworking student researchers of the Computational Speech Lab. Our work includes live linguistic data collection, corpus anlaysis, and machine learning for spoken language. For more information about what students are working on, please see our lab website.
Work with Me
SFU students interested in doing research with me should send an e-mail to sara_ng (at) sfu (dot) ca, or discuss in person. The kind of projects I am happy to advise on include but are not limited to:
- phonetic analysis
- speech classification
- error analysis of speech technology
- speech recognition
- phonological analysis
- pathological speech/hearing disorders
- corpus collection and analysis
- social media research
- natural language processing
Are you an undergraduate? You do not need to have a project in mind to get started doing research! Let's talk about projects you can volunteer your time with to gain research experience and meet other motivated students.
Research Tool Demos
English Word Class Generator (beta)
English Minimal Pair Generator (beta)
"Tap a rhythm or play a melody" (mTurk) [Example 1][Example 2]