Textual Representations Supported by Visual Similarity for Image Mining in Social Network Sites

• Grant number: 12/00005-0
• Support Opportunities: Scholarships in Brazil - Doctorate
• Start date: May 01, 2012
• End date: April 30, 2017
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Agma Juci Machado Traina
• Grantee: Alceu Ferraz Costa
• Host Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil
• Associated scholarship(s): 13/14040-4 - Graph mining in social media services, BE.EP.DR

Abstract

Due to its growing importance, social network sites have become increasingly popular for people to express feelings, communicate information and share content. A significant part of such contents are in the form of images and for this reason, social network sites are a valuable source of information from which to conduct image mining. Existing approaches for image mining on social network sites can be divided into two categories: those based on textual data associated with images and representations based on low-level visual characteristics. Such approaches have problems related to noise and semantic inconsistencies. To enable the image mining in social networking sites we propose in this doctoral project a new image representation approach: textual representation supported by visual similarity (TRVS). The TRVS can be seen as a bridge between low-level visual features and textual data associated with the images and will be used to improve data mining on social networking sites.

Academic Publications

(References retrieved automatically from State of São Paulo Research Institutions)

COSTA, Alceu Ferraz. Mining User Activity Data in Social Media Services. 2017. Doctoral Thesis - Universidade de São Paulo (USP). Instituto de Ciências Matemáticas e de Computação (ICMC/SB) São Carlos.