Please use this identifier to cite or link to this item: https://hdl.handle.net/10356/62705
Title: Efficient photo summarization from popular social networks
Authors: Chu, Xiaoqi
Keywords: DRNTU::Engineering::Electrical and electronic engineering
Issue Date: 2015
Abstract: In modern social networks, photo sharing is one of the major activities that gain rapid growing popularity in communication. These photos are scattered across many different social networks, such as Facebook, Twitter, Instagram, etc. It is difficult for people to gather the relevant photos under a specific topic. Photo summarization is an algorithm to fetch relevant photos from social networks based on metadata, comments, etc. In order to fulfill the requirement, the solution is capable of getting data from multiple platforms, handling data coming in high traffic, dealing with a great amount of data, and crawling data which is loaded by JavaScript asynchronously. The whole research involved a number of techniques, including scraping tool Scrapy, web automation browser driver Selenium, NoSQL database MongoDB, Tornado asynchronous web server, etc. Two ways of scraping information was analyzed and adopted. A chrome plugin was developed in order to increase the efficiency of scraping relevant data. A high-performance distribute-deployment server was fully implemented to store and dispatch data. This project topic was initiated by two real projects in IT industry. PhoneSoul PTE LTD and FOMO Digital PTE LTD used the result of this research in their products, both of which have been commercially used. By adopting the technology discussed in this report, PhoneSoul could summarize a great amount of photos related to a certain topic in order to do data mining for their product purpose, and FOMO Digital could get certain newly uploaded photos on social media platforms related to some topics in real time.
URI: http://hdl.handle.net/10356/62705
Rights: Nanyang Technological University
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:SCSE Student Reports (FYP/IA/PA/PI)

Files in This Item:
File Description SizeFormat 
FYP_SCE14293_Chu_Xiaoqi_reduced.pdf
  Restricted Access
2.22 MBAdobe PDFView/Open

Page view(s)

108
checked on Sep 27, 2020

Download(s)

18
checked on Sep 27, 2020

Google ScholarTM

Check

Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.