Microsoft Research

Incidents involved as both Developer and Deployer

Incident 121 Report
Common Biases of Vector Embeddings

2016-07-21

Researchers from Boston University and Microsoft Research, New England demonstrated gender bias in the most common techniques used to embed words for natural language processing (NLP).

Related Entities

Entity

Boston University

Incidents involved as both Developer and Deployer

Incident 12
1 Report
Common Biases of Vector Embeddings

Entity

Google

Incidents involved as Developer

Incident 12
1 Report
Common Biases of Vector Embeddings

Entity

Women

Incidents Harmed By

Incident 12
1 Report
Common Biases of Vector Embeddings

Entity

Minority Groups

Incidents Harmed By

Incident 12
1 Report
Common Biases of Vector Embeddings

Microsoft Research

Incidents involved as both Developer and Deployer

Incident 121 ReportCommon Biases of Vector Embeddings

Related Entities

Boston University

Incidents involved as both Developer and Deployer

Google

Incidents involved as Developer

Women

Incidents Harmed By

Minority Groups

Incidents Harmed By

Incident 121 Report
Common Biases of Vector Embeddings