Deep Language models for Visual analytics of diversity and inclusion

The project will build an Artificial Intelligence-powered Bias Detection tool to identify bias and discrimination faced by employees of a company. The tool will take in free-text content from employee surveys, internal reviews and social media, and produce a score indicating whether the author of the content is likely facing or exhibiting bias. Algorithms will be trained to recognize bias against women, ethnic & cultural minorities, LGBTQ2+ folks, individuals with disabilities, and other special groups. The tool will help companies in improving their diversity and inclusion practices, and investors in focussing their funds on companies with a strong diversity and inclusion record. The results of the project will help Diversio achieve and maintain world-class status in this emerging IT industry sector.

Faculty Supervisor:

Evangelos Milios

Student:

Maksym Taranukhin

Partner:

Diversio

Discipline:

Computer science

Sector:

Professional, scientific and technical services

University:

Dalhousie University