Ranking Instagram Preferences: Get to know your friends better through experimental mathematics Open Access

Choubal, Urshila (Spring 2022)

Permanent URL: https://etd.library.emory.edu/concern/etds/zw12z6651?locale=en


Ranking methods offer remarkable potential in creating and revamping recommendation systems. The task of suggesting more relevant and attractive content to users is directly benefited by improving ranking techniques. With graph ranking as the mathematical foundation on which recommendation systems are built, vertex prestige is a critical problem to be addressed. Several models exist that rank vertices in a graph. However, we explore the following methods: HITS, Dominant Eigenvector, and PageRank. We aim to emulate a recommendation system by first gathering primary data from Instagram by tracking the activity of nine participants on the app. With the help of the three ranking methods, we intend to provide our recommendation to the participants based on having accessed their past preferences. 

Table of Contents

1. Social Media and User Preferences

1.1 Recommendation Systems

1.2 Clustering and Ranking

1.3 Instagram: the photo-video sharing app

2. Graph Ranking: Theory and Algorithms

2.1 Preliminaries

2.2 Graph Theory

2.3 Ranking Methods Used in Our Experiment

3. An Instagram Experiment

3.1 Data Collection

3.2 Data Interpretation

3.3 Numerical Results

3.4 Discussion of Our Findings

4. Conclusions and Future Work

About this Honors Thesis

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
  • English
Research Field
Committee Chair / Thesis Advisor
Committee Members
Last modified

Primary PDF

Supplemental Files