Information Spread in Online Social Media
Abstract
The goal of this research was to study viral information spread occurring in social media sites. This phenomenon is referred to as information cascades. This research investigated properties of massive network of users of a social media site VK.com, studied networks of adviser-advisee relationships in mathematics, networks formed by word2vec word embeddings, and networks of causal relationships in the U.S. stock market. In addition this research investigated how deep neural networks can be applied for prediction of cascades in social media sites VK.com and Weibo, developed multitask machine learning methods for native language identification, and applied transfer learning for novel recommender system algorithms. Finally, this grant supported the investigation of how difference of convex functions (DC) programming may be used for solving network optimization problems. This project resulted in five published papers in peer-reviewed journals, five peer-reviewed conference proceedings, and three submitted papers.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 25, 2020
- Accession Number
- AD1108765
Entities
People
- Alexander Semenov
Organizations
- University of Jyväskylä