SXSW 2023

Positive Play - Toxicity Detection in Voice Chat

Description:

Bullying, gender discrimination, and other forms of verbal toxicity that frequently occur during multi-player game play cost the gaming industry billions of dollar each year in dropped subscriptions and lost customers... and it's directly counter to the inclusivity and accessibility inherent in gaming that we want to be promoting. It isn't just about the language, it's even more about tone and context. Eliminating toxicity to create safe, inclusive spaces for game play is a highly complex challenge, especially for multiplayer games that incorporate realtime voice chat - but it CAN be solved through the intelligent application of machine learning. Solutions must be flexible, allowing for different tolerances and languages, and enabling a range of remediation and escalation options.


Related Media

Additional Supporting Materials

Other Resources / Information


Takeaways

  1. How does toxicity show up in game play, and how much is it currently costing the gaming industry?
  2. What are effective approaches for predicting, identifying and responding to verbal in-game toxicity, and what are the most difficult challenges?
  3. How can machine learning and ensemble modeling be applied to create flexible, tunable anti-toxicity solutions?

Speakers


Organizer

Stephen Bradley, Managing Director, Deloitte Digital


Meta Information:

  • Event: SXSW
  • Format: Panel
  • Track: Games & Entertainment
  • Track 2
  • Level: Advanced


Add Comments

comments powered by Disqus

SXSW reserves the right to restrict access to or availability of comments related to PanelPicker proposals that it considers objectionable.