H o m e :: Publications :: Contact
R e s e a r c h I n t e r e s t s
My research focuses on the automatic processing of natural language.
Here are some topics of interest to me:
- Information extraction and classification
- Speech interfaces and multimodal interfaces
- Machine learning
- Information retrieval
Ubiquitous Computing
In addition to my work at Intel, I am a lecturer at the UC Berkeley School of Information, where I teach a class on Applied Natural Language Processing (fall 09).
Most recently, I developed a speech interface for mobile devices for a platform composition system and for the Digital Living Room (i.e. to talk with the TV) (photos of myself demonstrating the interfaces:
1 2).Platform Composition is a technique that integrates standard computing components to support effective collaborative work by wirelessly combining the most suitable set of resources available on nearby devices. Some papers describing my work in this area can be found here, here and here.
I am interested in natural and spoken language processing and systems research for mobile devices. I organized a workshop on this topic, Mobile Language Processing ACL workshop.
I published a book See also my publications.
M y b a c k g r o u n d
I got my PhD in 2005 from the iSchool, UC Berkeley, under the supervision of Marti Hearst.
My PhD thesis was on natural language processing for medical text, specifically, on the semantic interpretation of noun compounds, on the extraction of semantic entities such as treatment, disease, protein and the interpretation of the relationships between them. I developed machine learning algorithms to tackle these problems. [papers here here and here]
My thesis: Extraction of semantic relations from bioscience text.
In 1997-98, I worked as a visiting scholar at the Media Laboratory at MIT (with Sandy Pentland), where I was involved in a project using machine learning methods (Bayesian networks) for the modeling and recognition of human behavior; tasks included the classification of two-handed motions in the martial art ``Tai Chi'' and in American Sign Language, and the categorization of multiple-person interactions in a visual surveillance task. [papers here here and here].
Fashionable Wearable American Sign Language Recognition System: my participation at the MIT Media Lab Wearable Fashion Show: Beauty and the bits, what a wonderful combination.
I also worked at the University of Cambridge, England, as a research associate; the project involved the development of machine learning methods for the analysis of medical data. [paper]
And in the ancient past, I studied Physics at the University of Trieste, where I graduated with a thesis on Neural Networks. [paper]
H o m e :: Publications :: Contact