Detecting Hate Speech in online Media

We are sorry, this position has been filled.

This project will involve developing a classifier to automatically detect whether hate speech, bias or uncivil discourse exists in an online post. The project will focus on hate speech aimed at a particular community including one of the following: LGBQ, female, female journalists, blacks, ethnic group (e.g., latinos), religious minority (e.g., Muslim). The student will use an existing dataset and develop a classifier using a supervised approach using a standard toolkit such as SciKit Learn. The task will be to identify features that are important in developing an accurate classifier.

Lab: Natural Language Processing

Direct Supervisor: Kathleen McKeown

Position Dates: 6/1/2018 - 8/30/2018

Hours per Week: 20

Paid Position: Yes

Credit: No

Qualifications: Python

Eligibility: Sophomore; SEAS only

Kathleen McKeown, kathy@cs.columbia.edu