Spark! Alum, Where Are They Now? Jorge Nario
Most people dream of seeing the culmination of their work in a published paper. Spark! alum and Google Software Engineer Jorge Nario (CAS’20) made that a reality with his recent work, “Capturing Covertly Toxic Speech via Crowdsourcing.” As one of the brains behind the Google team who authored the paper, Nario worked to address the issue of “covert” or “veiled” toxicity, such as microaggressions or obfuscations in text.
How long were you involved in Spark and in what capacity?
Jorge: I started my involvement with Spark! my sophomore year. I worked as the Co-Chair for the Ignite Counsel. I also participated in the Fall 2019 Fellow program where my group TallTale won best in tech.
How did you get to where you are today?
Jorge: I believe internships and my independent research at BU was the biggest factor in getting me into A.I.
What is something you wish you knew as a student?
Jorge: Some things may just be out of your control. You will not ace every test and job interview, and that is okay. You can never get everything perfect; it’s best to prioritize learning, and the rest will work itself out in the longterm.
Please tell us about your published paper “Capturing Covertly Toxic Speech via Crowdsourcing,” it’s importance and the efforts that went into it.
Jorge: This paper offers a few major contributions to this space [text that hides malicious intent]. The first one is a covertly offensive crowdsourcing template that can be adapted for similar problems … The second contribution is a now publicly available Covert TensorFlow dataset (civil_comments/