– Case Study –
Online Gaming Platform Powers Down Inappropriate Content
Platform ensures chat logs are a safe haven for its virtual community
Challenge
A leading online gaming platform needed help reviewing vast amounts of chat logs line-by-line to determine if the content is appropriate for its all-ages audience.
Results
SOLUTION
Innodata developed a chatbot to automatically read and identify inappropriate text. We built a comprehensive taxonomy of inappropriate short phrases and idioms and then applied the corresponding corrections to them. Judgments were verified by SMEs. Further enhancement to the solution involved the addition of ML to analyze the chat logs and automatically classify the text.
IMPACT
The automated process for review and correction of chat logs ensures content is age-appropriate. The addition of ML further improved the accuracy and speed of flagging inappropriate content. The feed is used to continually train the ML model to better recognize inappropriate content. Its assessment is supplemented by an SME for QA.
Meet an Expert
Our Team of Data Experts
A team comprised of data experts with extensive experience in developing AI-based data solutions for clients. Book a time that works for you and let us help develop a custom solution for your unique needs.