In this project we  study automatic tag extraction from social media content for open domain image labeling. Our goal is to design and develop social media content analysis tools that will generate tags for visual content and capture user's interest and hobbies. We will explore the use of key words extraction and abstraction, n-grams models, ontology and semantic resources. Our focus is on building and testing of solution prototypes. 

Financing: Part of the Digile D2I research program, funded by Tekes, Arcada Research Foundation

Project partners: F-secure, PacketVideo

Duration: January 2014 -

Project manager: Shuhua Liu

Participating researchers: Shuhua Liu