Relational Social Media Search Engine

DSpace/Manakin Repository

Relational Social Media Search Engine

Show full item record

Title: Relational Social Media Search Engine
Author(s):
Banipal, Indervir;
0000-0003-4151-6807
Advisor: Khan, Dr. Latifur
Date Created: 2016-12
Format: Thesis
Keywords: Show Keywords
Abstract: Given a collection of names of some people who are related to each other by a common relation, our goal is to find their profiles from social networks/media. For instance, they might belong to a particular class from a particular university. The problem poses severe challenges as there might be numerous people present in social media with the same name. The goal is to propose novel algorithms which will help us extracting the correct set of people from enormous possibilities. Social networks have become a primary source of communication and the number one choice for people to stay in touch with each other. Apart from the use of social networks in private life, people rely heavily on the social networks for their professional lives as well. We exploit this fact and use social networks to find the relevant people we are looking for. Facebook [11], LinkedIn [12] and Twitter [13] are the most widely used social networks, for staying in touch with others privately, as well as professionally. Our approaches are applicable to any of these major social networks although we focus on Facebook. We go through the chosen social network and search for the names in the collection. Every name in the input collection should be searched in the social network and its results are stored. After the searching through the social network, we will have a list of search results for each of the names in the collection. For each user, the social network has suggested multiple possible social network pro les which may or may not contain the target user which we are v looking for. We record all such search results for each user and this serves as our baseline. The goal is to improve this baseline/non-veri ed list by augmenting with semi-automated method to prune some irrelevant pro les and nd the relevant one. We will discuss about social media crawler which we developed for extracting the information from the social media. We will also talk about the algorithms which we designed and implemented for showing the relevant pro les to the user. We will also talk about the human intervention in the form of feedback and nally, we present a complete generic framework for solving this problem.
Degree Name: MSCS
Degree Level: Masters
Persistent Link: http://hdl.handle.net/10735.1/5357
Type : text
Degree Program: Computer Science

Files in this item

Files Size Format View
BANIPAL-THESIS-2016.pdf 3.949Mb PDF View/Open

This item appears in the following Collection(s)


Show full item record