activity
20142026
most citedBias and Discrimination in AI: a cross-disciplinary perspective

306 citations · 851 across the 25 of their papers we have counts for

collaborators
Showing 2020Show all

5 papers · 1 filter

cs.CR2020★ 7 cited

The Challenges with Internet of Things for Business

Ievgeniia Kuzminykh, Bogdan Ghita, Jose M. Such

Many companies consider IoT as a central element for increasing competitiveness. Despite the growing number of cyberattacks on IoT devices and the importance of IoT security, no st…

cs.CL2020★ 3 cited

Discovering and Interpreting Biased Concepts in Online Communities

Xavier Ferrer-Aran, Tom van Nuenen, Natalia Criado +1

Language carries implicit human biases, functioning both as a reflection and a perpetuation of stereotypes that people carry with them. Recently, ML-based NLP methods such as word…

cs.CL2020★ 6 cited

Discovering and Categorising Language Biases in Reddit

Xavier Ferrer, Tom van Nuenen, Jose M. Such +1

We present a data-driven approach using word embeddings to discover and categorise language biases on the discussion platform Reddit. As spaces for isolated user communities, platf…

cs.CY2020★ 306 cited

Bias and Discrimination in AI: a cross-disciplinary perspective

Xavier Ferrer, Tom van Nuenen, Jose M. Such +2

With the widespread and pervasive use of Artificial Intelligence (AI) for automated decision-making systems, AI bias is becoming more apparent and problematic. One of its negative…

cs.AI2020★ 6 cited

A Normative approach to Attest Digital Discrimination

Natalia Criado, Xavier Ferrer, Jose M. Such

Digital discrimination is a form of discrimination whereby users are automatically treated unfairly, unethically or just differently based on their personal data by a machine learn…