Jigsaw Multilingual Toxic Comment Classification

Nancy J. Delong

It only requires a single harmful comment to sour an online discussion. The Conversation AI workforce, a investigate initiative launched by Jigsaw and Google, builds know-how to guard voices in the conversation. A most important place of focus is machine learning models that can identify toxicity in online discussions, wherever toxicity is outlined as anything impolite, disrespectful, or if not possible to make an individual depart a discussion. If these harmful contributions can be discovered, we could have a safer, more collaborative world-wide-web.

Jigsaw’s API, Perspective, serves toxicity models, and others in a expanding established of languages. More than the past year, the area has noticed outstanding multilingual abilities from the most up-to-date model innovations, such as handful of- and zero-shot learning.

Submission to this Problem should be obtained by eleven:fifty nine PM UTC, June 15, 2020.

Source: Kaggle


Next Post

Visual Studio Online becomes Visual Studio Codespaces

Microsoft has altered the title of Visual Studio Online to Visual Studio Codespaces, to bring the title a lot more in line with the idea that the offering is “more than just an ‘editor in a browser,’” the firm wrote in a web site write-up. Microsoft launched the title change on April […]