Dr. Fawad Hussain
University of Birmingham Dubai, UAE
Can artificial intelligence end artificial trends on twitter?
Syed Fawad Hussain obtained his MS degree in Computer Science from the prestigious Pierre and Marie Curie University (now Paris-Sorbonne University), Paris, France, and a Ph.D. in Computer Science from the University of Grenoble, France. During his illustrious career, he has worked in diverse areas. He was associated with ERIC lab, Lyon where he helped develop a personalized healthcare warehouse for the French National Football team. Afterwards, he worked on project FRAGRANCE funded partly by the French National Research for Scientific Research (CNRS) and Xerox Research Europe on defining similarity measures in linked graphs with application to text mining and bioinformatics. Dr. Fawad Hussain is an HEC approved PhD supervisor, a professional member of the ACM, a senior member of IEEE, co-chair of the HEC National Curriculum Review Committee for Computer Science, technical panel of the Pakistan Science Foundation (PSF). He is also the chief judge at the International Collegiate Programming Competition (ICPC) for the Asia-West (Topi) region. He is also recipient of multiple research and teaching awards. He was awarded the Best University Teacher Award (BUTA) by the Higher Education Commission of Pakistan, and the Best Research Award at the GIK Institute, for the year 2015. He has been selected as a Distinguished International Scholar by COMSTECH under the Organization of Islamic Countries (OIC). He is a recipient of the Google/IBM grant for doctoral forum at the SIAM data mining conference (SDM).
Social media is a powerful tool today that can help shape opinions, create movements, promote a cause, etc. but also cause mass hysteria and spread hate speech, false news, etc. Twitter is one such powerful tool with several hundred million users and around 500 million tweets per day. Unfortunately, however, not all users have positive intentions and not all tweets are harmless. Many users, even backed by strong groups, exploit this media for promoting false or fake news. While Twitter regulates users and tweets, this hard to do in real or near real time. Hence, it is possible for a group of users to generate an artificial trend which can potentially influence other users and cause much harm before it is detected. This talk examines the use of automated algorithms using feature extraction, feature selection, link analysis, and machine learning algorithms to detect a fake (artificial) trend in near real-time. A case study by the speaker is presented that can be used to successfully detect fake trends and the user profiles behind it in near real time with high accuracy.