Angeliki Metallinou

Welcome to my website! 

I am currently working as a speech scientist at Amazon. I am working on developing speech and language technologies for Amazon Alexa and products including the Amazon Echo and others. My research interests include natural language understanding, machine learning, deep learning, dialog systems, affective computing, automatic speech recognition, education and healthcare applications. 

Before that, I was employed as a research scientist at Pearson Knowledge Technologies, part of PearsonAs part of my work there, I worked on the Pearson Versant spoken test, an automatically graded test that evaluates speaking proficiency (fluency, pronunciation, vocabulary, etc). I focused on improving test accuracy using machine learning and deep learning. In addition, I worked on automatic grading of essays and short answer questions, using machine learning and natural language processing.

I am also interested in the area of affective computing, that was the focus of my PhD research at University of Southern California (USC), at the Signal Analysis and Interpretation Lab (SAIL). During my PhD, I worked on emotion recognition from speech and facial/body expressions during dyadic interactions. I also explored methodologies for making use of context and modeling emotion evolution in conversations. My work was also applied in healthcare, including autism research, for studying non-typical emotional expressions. 

As a PhD student, I have also interned at Microsoft Research, where I focused on machine learning approaches, particularly discriminative models, for dialog state tracking for spoken dialog systems.

For more information about my work, please see my publications page. You can also contact me at ametallinou@gmail.com