Table of contents
1. Other pages in this release
Automated classification of web-scraped clothing data in consumer price statistics
Using statistical distributions to estimate weights for web-scraped price quotes in consumer price statistics
This is the first in a series of biannual articles that aim to update users on our research to modernise the measurement of consumer price inflation in the UK.
Our research programme involves the investigation of both new data sources and new methods. Our first notable milestone is in Quarter 1 (Jan to Mar) 2023, when we will begin the incorporation of new data and methods into our headline measures of consumer price statistics for certain targeted categories. Users can read about our plans and timeline for implementation in the article Introducing alternative data sources into consumer price statistics.
In this biannual series, we focus on the research that is necessary for us to automate the production of these data sources, while maintaining the quality and integrity of our consumer price statistics. In the first article, we look at how we can automatically classify millions of web-scraped clothing data to the consumer price statistics hierarchy, how we choose the most appropriate price index number methods for use with alternative data sources, and a method of approximating product-level sales values to ensure products that we include in our consumer price indices are representative of consumer spending. A full research program is outlined in our Consumer prices development work plan.
The research presented in this article should not be considered as final but should begin to demonstrate how some traditional methods will need to be transformed in order to process the magnitude of information that is becoming available to us. All results and statistics within this article should therefore be treated as experimental.
During this period of research, we will continue to work with our Advisory Panels for Consumer Prices (Technical) and (Stakeholder) and broader users. We plan to have made final decisions on methods to be used in these targeted categories by December 2021, including analyses of their potential impact. Users will be provided the opportunity to comment on the new methods and their impact in response to our research articles and throughout 2022 as we publish an experimental series including alternative data sources in parallel to our official measures. We intend to initialise their incorporation into our headline consumer price statistics in Quarter 1 2023.
We welcome feedback on any of the research completed to date to email@example.com, which may help shape the ongoing transformation of consumer price statistics to include new data and methods.Back to table of contents
Contact details for this Article
Telephone: +44 (0)1633 456900