During last years, several studies focused on the predictive capability of web data to forecast statistical indicators. Google Trends is a free web tool that quantify search-term volume on the search engine. The aim of this work is to forecast the household expenditures for consumption in Italy, using Google Trends related to particular expenditure keywords. Several prediction models have been tested, also including relevant leading indicators correlated to the household expenditures behavior, on a time series survey data from 2004 to 2016. The ARIMA model has been chosen and models with different lag structures have been tested. The model comparison shows that the results including Google data outperform those of both benchmark and augmented models.
GOOGLE TRENDS FOR NOWCASTING QUARTERLY HOUSEHOLD CONSUMPTION EXPENDITURE / Fasulo, Andrea; Guandalini, Alessio; Terribili, Marco D.. - In: RIVISTA ITALIANA DI ECONOMIA, DEMOGRAFIA E STATISTICA. - ISSN 0035-6832. - (2017).
GOOGLE TRENDS FOR NOWCASTING QUARTERLY HOUSEHOLD CONSUMPTION EXPENDITURE
Alessio Guandalini
;Marco D. Terribili
2017
Abstract
During last years, several studies focused on the predictive capability of web data to forecast statistical indicators. Google Trends is a free web tool that quantify search-term volume on the search engine. The aim of this work is to forecast the household expenditures for consumption in Italy, using Google Trends related to particular expenditure keywords. Several prediction models have been tested, also including relevant leading indicators correlated to the household expenditures behavior, on a time series survey data from 2004 to 2016. The ARIMA model has been chosen and models with different lag structures have been tested. The model comparison shows that the results including Google data outperform those of both benchmark and augmented models.File | Dimensione | Formato | |
---|---|---|---|
2017LXXI_N4_RIEDS_01_Fasulo_Guandalini_Terribili (1).pdf
accesso aperto
Tipologia:
Versione editoriale (versione pubblicata con il layout dell'editore)
Licenza:
Tutti i diritti riservati (All rights reserved)
Dimensione
463.24 kB
Formato
Adobe PDF
|
463.24 kB | Adobe PDF |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.