Examining the Impact of the Internet News Environment on the Spread of Articles on Twitter

dc.contributor.advisor Zhang, June
dc.contributor.author Rodriguez, Michael
dc.contributor.department Electrical Engineering
dc.date.accessioned 2020-02-20T18:08:47Z
dc.date.available 2020-02-20T18:08:47Z
dc.date.issued 2019
dc.description.degree M.S.
dc.identifier.uri http://hdl.handle.net/10125/66241
dc.subject Electrical engineering
dc.subject Information Propgation
dc.subject Time-Series Prediction
dc.title Examining the Impact of the Internet News Environment on the Spread of Articles on Twitter
dc.type Thesis
dcterms.abstract Time-Series prediction problems have proven to be a challenge in nearly every domain, especially domains that do not necessarily exhibit periodic trends, or may be influenced by outside actions. One such domain is the activity of users on a social network. The prediction task has been studied in the past, where methods included; individual article features and short term activity [1], social media activity alone [2] and probabilistic methods based on early social media activity [3][4]. In this work we will present a modification to the previous methods, by including previous periods frequency count of specific feature as a feature for prediction of spread of an article on Twitter. It is based on the biological theory of frequency-dependent selection. An empirical study of the impact of the features was conducted, and there appears to be no noticeable improvement when we consider the frequency of article features noted on the Internet to the prediction of the spread of those articles on Twitter. This suggests a weak connection between the two networks. The modification may work better if the frequency of articles on Twitter itself were considered.
dcterms.extent 65 pages
dcterms.language eng
dcterms.publisher University of Hawaiʻi at Mānoa
dcterms.rights All UHM dissertations and theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission from the copyright owner.
dcterms.type Text
local.identifier.alturi http://dissertations.umi.com/hawii:10499
Files
Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
Rodriguez_hawii_0085O_10499.pdf
Size:
3.49 MB
Format:
Adobe Portable Document Format
Description: