Please use this identifier to cite or link to this item:
Title: A late fusion multimodal sentiment analysis model and hashtags summarizer with interactive visualisation for geotagged Instagram posts
Authors: Tan, Calvin Sin Nian
Keywords: DRNTU::Engineering::Computer science and engineering
Issue Date: 2019
Abstract: The proliferation of social media platforms in recent years has allowed people to voice their opinions in quick and easy ways. The advent of smartphones further enabled users to share information on-the-go. As a result, massive amount of data can be harvested and be analysed to crowdsource public sentiment on certain topics of interest. These insights can be useful for aiding decisions in business, social, and political contexts. Platforms like Instagram allows businesses to create geographical identification metadata or hashtags to identify with their businesses. This metadata can then be repeatedly used by users to associate their media content with these geotags or hashtags. Instagram Scraper, a command-line application, enables the extraction of such geotagged or hashtagged media contents. The extracted media can be used for textual and visual analyses. This project aims to create an interactive dashboard to enable users to glance at the topic or entity of interest. Charts reflecting emotions, sentiments, and topics allow users to crowdsource opinions and attitudes, helping them to make more informed decisions. Visual sentiment analysis also breaks free of traditional textual sentiment analysis which can be misleading because captions used in social media are sometimes irrelevant to the images being posted. Finally, this paper also proposes a hybrid sentiment analysis model that integrates the various modes of sentiment analysis to give a holistic overview of sentiment scoring relevant to the image sharing social media platforms.
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 
  Restricted Access
1.93 MBAdobe PDFView/Open

Google ScholarTM


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