TY - GEN
T1 - Utilization of Artificial Intelligence for Social Media and Gaming Moderation
AU - Saleous, Heba
AU - Gergely, Marton
AU - Shuaib, Khaled
N1 - Publisher Copyright:
© 2023 IEEE.
PY - 2023
Y1 - 2023
N2 - As the world continues to evolve, technology has proven to be a necessity in the lives of everyone. Evolving beyond professional use, cyberspace is now populated by online communities being used for communication, learning, and entertainment. However, the increased online presence exposes users to a variety of cultures, personalities, and levels of maturity. Some may also seek to cause harm to others through cyberbullying or may display toxic behaviors. This research aims to tackle the growing problem of toxicity and harassment in online environments. The proposed solution will utilize Artificial Intelligence (AI), and more specifically Natural Language Processing (NLP), to moderate communication and detect malicious language and behavior. The efforts shared in this paper specifically present a work-in-progress. For the time being, two models have been tested with a single dataset from Twitter: A Convolutional Neural Network (CNN) and a Recurrent Neural Network (RNN). The results of experimentation show a promising start for the use of NLP in moderation with an 83% accuracy using an RNN.
AB - As the world continues to evolve, technology has proven to be a necessity in the lives of everyone. Evolving beyond professional use, cyberspace is now populated by online communities being used for communication, learning, and entertainment. However, the increased online presence exposes users to a variety of cultures, personalities, and levels of maturity. Some may also seek to cause harm to others through cyberbullying or may display toxic behaviors. This research aims to tackle the growing problem of toxicity and harassment in online environments. The proposed solution will utilize Artificial Intelligence (AI), and more specifically Natural Language Processing (NLP), to moderate communication and detect malicious language and behavior. The efforts shared in this paper specifically present a work-in-progress. For the time being, two models have been tested with a single dataset from Twitter: A Convolutional Neural Network (CNN) and a Recurrent Neural Network (RNN). The results of experimentation show a promising start for the use of NLP in moderation with an 83% accuracy using an RNN.
KW - NLP
KW - Online Harassment
KW - Sentiment Analysis
KW - User Moderation
UR - http://www.scopus.com/inward/record.url?scp=85182918278&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85182918278&partnerID=8YFLogxK
U2 - 10.1109/IIT59782.2023.10366468
DO - 10.1109/IIT59782.2023.10366468
M3 - Conference contribution
AN - SCOPUS:85182918278
T3 - 2023 15th International Conference on Innovations in Information Technology, IIT 2023
SP - 246
EP - 251
BT - 2023 15th International Conference on Innovations in Information Technology, IIT 2023
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 15th International Conference on Innovations in Information Technology, IIT 2023
Y2 - 14 November 2023 through 15 November 2023
ER -