12:30 - 13:30
Tu-SPOT02
Chair:
Albrecht Wirthmann (Eurostat, Luxembourg)
Computation of consumer spatial price indexes over time using Natural Language Processing and web scraping techniques
Tiziana Laureti, (Email) 1, Ilaria Benedetti, (Email) 1, Luigi Palumbo, (Email) 1, Brandon Rose, (Email) 2
1 Università degli Studi della Tuscia
2 Starsift LLC

The use of Big Data in official statistics has become a major topic at international level. Among all the possible types of Big Data sources, the “Internet as a data source” is a popular one, and consumer price statistics may benefit from leveraging this source.

In this paper we used web scraped data to compile spatial price indexes for 93 categories according to COICOP level 4 and track their evolution over time in 11 US cities reported in our dataset. Given space limitation, we only report results for 2 of those categories: Apples and Chocolate.