The power of twitter on predicting box office revenues

Jooyoung Jeon, Patrick McSharry

Research output: Working paper

Abstract

Over the last few years there has been an extraordinary surge of social networking and microblogging services. Twitter is a social network that focuses on social and news media. The Twitter data stream allows access to tweets, timestamps and locations of users. This enables us to capture the trends and patterns of rapidly evolving worldwide events. We use the Twitter data stream for the prediction of consumer preferences in the movie industry and estimate how successful the movie will be in the first and second weekends since its release date. The study provides evidence to suggest that frequencies of contemporaneous tweets and a consensus measure of public sentiment are useful for predicting box-office revenues, implying that any publicity is good publicity in word-of-mouth (WOM) and online viral marketing. Sentiment analysis based on tweets suggests that more extreme sentiment has more impact, and that the more negative the tweets about a movie are, the higher its revenue will be, in contrast with the classic theory of diffusion in news media.
LanguageEnglish
Pages2-22
Number of pages21
Publication statusUnpublished - 7 May 2012

Fingerprint

Marketing
Industry

Keywords

  • social media
  • prediction
  • box office revenue
  • regression

Cite this

Jeon, J., & McSharry, P. (2012). The power of twitter on predicting box office revenues. (pp. 2-22).
Jeon, Jooyoung ; McSharry, Patrick. / The power of twitter on predicting box office revenues. 2012. pp. 2-22
@techreport{9a9c4c749b6b425587a7192eb928b019,
title = "The power of twitter on predicting box office revenues",
abstract = "Over the last few years there has been an extraordinary surge of social networking and microblogging services. Twitter is a social network that focuses on social and news media. The Twitter data stream allows access to tweets, timestamps and locations of users. This enables us to capture the trends and patterns of rapidly evolving worldwide events. We use the Twitter data stream for the prediction of consumer preferences in the movie industry and estimate how successful the movie will be in the first and second weekends since its release date. The study provides evidence to suggest that frequencies of contemporaneous tweets and a consensus measure of public sentiment are useful for predicting box-office revenues, implying that any publicity is good publicity in word-of-mouth (WOM) and online viral marketing. Sentiment analysis based on tweets suggests that more extreme sentiment has more impact, and that the more negative the tweets about a movie are, the higher its revenue will be, in contrast with the classic theory of diffusion in news media.",
keywords = "social media, prediction, box office revenue, regression",
author = "Jooyoung Jeon and Patrick McSharry",
year = "2012",
month = "5",
day = "7",
language = "English",
pages = "2--22",
type = "WorkingPaper",

}

Jeon, J & McSharry, P 2012 'The power of twitter on predicting box office revenues' pp. 2-22.

The power of twitter on predicting box office revenues. / Jeon, Jooyoung; McSharry, Patrick.

2012. p. 2-22.

Research output: Working paper

TY - UNPB

T1 - The power of twitter on predicting box office revenues

AU - Jeon, Jooyoung

AU - McSharry, Patrick

PY - 2012/5/7

Y1 - 2012/5/7

N2 - Over the last few years there has been an extraordinary surge of social networking and microblogging services. Twitter is a social network that focuses on social and news media. The Twitter data stream allows access to tweets, timestamps and locations of users. This enables us to capture the trends and patterns of rapidly evolving worldwide events. We use the Twitter data stream for the prediction of consumer preferences in the movie industry and estimate how successful the movie will be in the first and second weekends since its release date. The study provides evidence to suggest that frequencies of contemporaneous tweets and a consensus measure of public sentiment are useful for predicting box-office revenues, implying that any publicity is good publicity in word-of-mouth (WOM) and online viral marketing. Sentiment analysis based on tweets suggests that more extreme sentiment has more impact, and that the more negative the tweets about a movie are, the higher its revenue will be, in contrast with the classic theory of diffusion in news media.

AB - Over the last few years there has been an extraordinary surge of social networking and microblogging services. Twitter is a social network that focuses on social and news media. The Twitter data stream allows access to tweets, timestamps and locations of users. This enables us to capture the trends and patterns of rapidly evolving worldwide events. We use the Twitter data stream for the prediction of consumer preferences in the movie industry and estimate how successful the movie will be in the first and second weekends since its release date. The study provides evidence to suggest that frequencies of contemporaneous tweets and a consensus measure of public sentiment are useful for predicting box-office revenues, implying that any publicity is good publicity in word-of-mouth (WOM) and online viral marketing. Sentiment analysis based on tweets suggests that more extreme sentiment has more impact, and that the more negative the tweets about a movie are, the higher its revenue will be, in contrast with the classic theory of diffusion in news media.

KW - social media

KW - prediction

KW - box office revenue

KW - regression

M3 - Working paper

SP - 2

EP - 22

BT - The power of twitter on predicting box office revenues

ER -

Jeon J, McSharry P. The power of twitter on predicting box office revenues. 2012 May 7, p. 2-22.