close

Have questions, suggestions, or concerns?

Program Directors:

Mailing Address:

Siebel Scholars Foundation
1300 Seaport Blvd., Suite 400
Redwood City, CA 94063

Phone:

(650) 299-5260

2008 - Shilpa Arora

Shilpa teaches computers how to efficiently extract information and sentiment from text.

Most recently, Shilpa has devised a method for training computers how to quickly identify the difference between negative and positive comments found in online reviews, social networking sites, and blogs.  This process, known as opinion mining, allows marketers to scour the web and uncover their consumers’ satisfaction with a particular feature, product, or movie. Shilpa’s biggest challenge is teaching the computer why a review is negative, not just that it is negative.  In her research, she strives for an approach that automatically recognizes linguistic patterns associated with opinions.

In addition to advertising, opinion mining has many potential applications, including several in the healthcare industry.  For example, it can be used to detect uncertain and negative assertions in most BioMedical Text Mining.  Also, Shilpa thinks opinion mining approaches could be used to identify patterns more quickly and accurately in psychiatric patients’ writings than today – patterns that can be used to automatically make at least a preliminary assessment of the patient’s mental health.

In teaching the computer to extract new patterns, Shilpa’s overall goal is to reduce the human annotator effort through the selective querying of desired information.

Shilpa is currently a Ph.D. candidate at the Languages Technologies Institute at Carnegie Melllon and has many publications, including a paper called, “Active Learning in a Multi-Annotator Environment,” the result of a class module that she designed and taught at Carnegie Mellon. She loves teaching, reading, and dancing Kathak, among other dance styles.