Document worth reading: “Social Media-based User Embedding: A Literature Review”

Automated illustration finding out is behind many newest success tales in machine finding out. It is often used to change data found from an enormous dataset (e.g., raw textual content material) to duties for which solely a small number of teaching examples may be discovered. In this paper, we analysis newest advance in finding out to characterize social media clients in low-dimensional embeddings. The know-how is crucial for creating extreme effectivity social media-based human traits and conduct fashions as a result of the ground actuality for assessing latent human traits and conduct is often expensive to build up at an enormous scale. In this survey, we analysis typical methods for finding out a unified individual embeddings from heterogeneous individual data (e.g., combines social media texts with images to be taught a unified individual illustration). Finally we stage out some current factors and future directions. Social Media-based User Embedding: A Literature Review