BIAS LANGUAGE SCANNER



Web Design & Development, React & Python


DEI-SJ Scanner — stands for Diversity, Equity, Inclusion and Social Justice Scanner, is a tool that the IT Department at The New School has initiated to detect the use of inappropriate, offensive, gender-coded, or otherwise objectionable language in the university’s daily communications — especially in job postings — that might lead to potential diversity, equity, and inclusion issues.

During my Full Stack Developer Internship with The New School’s IT, I transformed the prototype into a functional web application, after strengthening its core algorithm with multi-disciplinary efforts and creating a webapp from scratch. 
 



How it works:





The algorithm of the DEI-SJ Scanner has aggregated multi-disciplinary efforts, including academic research on gender studies and labor economics, open-source tools and natural language processing, as the following shows: 


Academic Research

Evidence That Gendered Wording in Job Advertisements Exists and Sustains Gender Inequality, Danielle Gaucher, Justin Friesen, Aaron C. Kay, 2011 

Older Workers Need Not Apply? Ageist Language In Job Ads And Age Discrimination In Hiring, National Bureau Of Economic Research, 2019 


Tools

Gender Decoder - Finding subtle bias in job ads
Alex - A JS package for catching insensitive, inconsiderate writing 

SpaCy - Industrial-Strength Natural Language Processing