Mining for Signals of Future Consumer Expenditure on Twitter and Google Trends

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Consumer expenditure constitutes the largest component of Gross Domestic Product in developed countries, and forecasts of consumer spending are therefore an important tool that governments and central bank use in their policy-making. In this paper we examine methods to forecast consumer spending from user-generated content, such as search engine queries and social media data, which hold the promise to produce forecasts much more efficiently than traditional surveys. Specifically, the aim of the paper is to study the relative utility of evidence about purchase intentions found in Google Trends versus those found in Twitter posts, for the problem of forecasting consumer expenditure. Our main findings are that, firstly, the Google Trends indicators and indicators extracted from Twitter are both beneficial for the forecasts: adding them as exogenous variables into regression model produces improvements on the pure AR baseline, consistently across all the forecast horizons. Secondly, we find that the Google Trends variables seem to be more useful predictors than the semantic variables extracted from Twitter posts, the differences in performance are significant, but not very large.
Original languageEnglish
Title of host publicationProceedings of 2nd International Conference on Advanced Reserach Methods and Analytics (Internet and Big Data in Economics and Social Sciences)
Place of PublicationValencia, Spain
Pages157-165
Number of pages8
ISBN (Electronic)9788490486894
DOIs
Publication statusPublished - 13 Jul 2018
Event2nd International Conference on Advanced Research Methods and Analytics (CARMA 2018) - Valencia, Spain
Duration: 12 Jul 201813 Jul 2018

Conference

Conference2nd International Conference on Advanced Research Methods and Analytics (CARMA 2018)
CountrySpain
CityValencia
Period12/07/1813/07/18

Fingerprint

Google
Twitter
Consumer expenditure
Consumer spending
Central bank
Regression model
Gross domestic product
Social media
Predictors
Purchase intention
Exogenous variables
User-generated content
Search engine
Query
Developed countries
Government
Forecast horizon
Policy making

Bibliographical note

This work is licensed under a Creative Commons License CC BY-NC-ND 4.0

Cite this

Pekar, V. (2018). Mining for Signals of Future Consumer Expenditure on Twitter and Google Trends. In Proceedings of 2nd International Conference on Advanced Reserach Methods and Analytics (Internet and Big Data in Economics and Social Sciences) (pp. 157-165). Valencia, Spain. https://doi.org/10.4995/CARMA2018.2018.8337
Pekar, Viktor. / Mining for Signals of Future Consumer Expenditure on Twitter and Google Trends. Proceedings of 2nd International Conference on Advanced Reserach Methods and Analytics (Internet and Big Data in Economics and Social Sciences). Valencia, Spain, 2018. pp. 157-165
@inproceedings{af35b072d8ca47f2a113d6ec415014de,
title = "Mining for Signals of Future Consumer Expenditure on Twitter and Google Trends",
abstract = "Consumer expenditure constitutes the largest component of Gross Domestic Product in developed countries, and forecasts of consumer spending are therefore an important tool that governments and central bank use in their policy-making. In this paper we examine methods to forecast consumer spending from user-generated content, such as search engine queries and social media data, which hold the promise to produce forecasts much more efficiently than traditional surveys. Specifically, the aim of the paper is to study the relative utility of evidence about purchase intentions found in Google Trends versus those found in Twitter posts, for the problem of forecasting consumer expenditure. Our main findings are that, firstly, the Google Trends indicators and indicators extracted from Twitter are both beneficial for the forecasts: adding them as exogenous variables into regression model produces improvements on the pure AR baseline, consistently across all the forecast horizons. Secondly, we find that the Google Trends variables seem to be more useful predictors than the semantic variables extracted from Twitter posts, the differences in performance are significant, but not very large.",
author = "Viktor Pekar",
note = "This work is licensed under a Creative Commons License CC BY-NC-ND 4.0",
year = "2018",
month = "7",
day = "13",
doi = "10.4995/CARMA2018.2018.8337",
language = "English",
pages = "157--165",
booktitle = "Proceedings of 2nd International Conference on Advanced Reserach Methods and Analytics (Internet and Big Data in Economics and Social Sciences)",

}

Pekar, V 2018, Mining for Signals of Future Consumer Expenditure on Twitter and Google Trends. in Proceedings of 2nd International Conference on Advanced Reserach Methods and Analytics (Internet and Big Data in Economics and Social Sciences). Valencia, Spain, pp. 157-165, 2nd International Conference on Advanced Research Methods and Analytics (CARMA 2018), Valencia, Spain, 12/07/18. https://doi.org/10.4995/CARMA2018.2018.8337

Mining for Signals of Future Consumer Expenditure on Twitter and Google Trends. / Pekar, Viktor.

Proceedings of 2nd International Conference on Advanced Reserach Methods and Analytics (Internet and Big Data in Economics and Social Sciences). Valencia, Spain, 2018. p. 157-165.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Mining for Signals of Future Consumer Expenditure on Twitter and Google Trends

AU - Pekar, Viktor

N1 - This work is licensed under a Creative Commons License CC BY-NC-ND 4.0

PY - 2018/7/13

Y1 - 2018/7/13

N2 - Consumer expenditure constitutes the largest component of Gross Domestic Product in developed countries, and forecasts of consumer spending are therefore an important tool that governments and central bank use in their policy-making. In this paper we examine methods to forecast consumer spending from user-generated content, such as search engine queries and social media data, which hold the promise to produce forecasts much more efficiently than traditional surveys. Specifically, the aim of the paper is to study the relative utility of evidence about purchase intentions found in Google Trends versus those found in Twitter posts, for the problem of forecasting consumer expenditure. Our main findings are that, firstly, the Google Trends indicators and indicators extracted from Twitter are both beneficial for the forecasts: adding them as exogenous variables into regression model produces improvements on the pure AR baseline, consistently across all the forecast horizons. Secondly, we find that the Google Trends variables seem to be more useful predictors than the semantic variables extracted from Twitter posts, the differences in performance are significant, but not very large.

AB - Consumer expenditure constitutes the largest component of Gross Domestic Product in developed countries, and forecasts of consumer spending are therefore an important tool that governments and central bank use in their policy-making. In this paper we examine methods to forecast consumer spending from user-generated content, such as search engine queries and social media data, which hold the promise to produce forecasts much more efficiently than traditional surveys. Specifically, the aim of the paper is to study the relative utility of evidence about purchase intentions found in Google Trends versus those found in Twitter posts, for the problem of forecasting consumer expenditure. Our main findings are that, firstly, the Google Trends indicators and indicators extracted from Twitter are both beneficial for the forecasts: adding them as exogenous variables into regression model produces improvements on the pure AR baseline, consistently across all the forecast horizons. Secondly, we find that the Google Trends variables seem to be more useful predictors than the semantic variables extracted from Twitter posts, the differences in performance are significant, but not very large.

UR - http://hdl.handle.net/10251/112101

UR - http://ocs.editorial.upv.es/index.php/CARMA/CARMA2018/paper/viewFile/8337/4286

U2 - 10.4995/CARMA2018.2018.8337

DO - 10.4995/CARMA2018.2018.8337

M3 - Conference contribution

SP - 157

EP - 165

BT - Proceedings of 2nd International Conference on Advanced Reserach Methods and Analytics (Internet and Big Data in Economics and Social Sciences)

CY - Valencia, Spain

ER -

Pekar V. Mining for Signals of Future Consumer Expenditure on Twitter and Google Trends. In Proceedings of 2nd International Conference on Advanced Reserach Methods and Analytics (Internet and Big Data in Economics and Social Sciences). Valencia, Spain. 2018. p. 157-165 https://doi.org/10.4995/CARMA2018.2018.8337