Together with Stichting Alexander, an expert in participatory research for young people, we are participating in a Small Business Innovation Research (SBIR) program of the Dutch government. Namely, in the program that focuses on improving inclusivity and empowerment by the use of Artificial Intelligence (AI). First we conducted a feasibility study, which we explained here. We will move forward in to create a working prototype of our idea for a more inclusive vacancy market. 

The Challenge: Not-inclusive Vacancies

Our feasibility study showed that miscommunication between employers and young job seekers is a prime cause of youth unemployment. This is an urgent problem as during the Covid crisis in 2020, we have seen youth unemployment rise by over 30% compared to 2019. Although there are enough open vacancies, employers struggle to find appropriate candidates for these vacancies. The reason for this is that currently, vacancies are often (unintentionally) not inclusive. They deter youth with a migration background, specific education level, gender, sexual orientation, or physical limitation, while they are appropriate candidates. It is important that we bridge this gap in miscommunication.

The Approach: Developing an AI Prototype with Partners

During the feasibility study, we held multiple youth panels to investigate which aspects of a vacancy can be perceived as non-inclusive. Based on this study, we have identified the five most important themes for young people:

(1) exclusion based on sex, age, or origin

(2) possibility of the employee to learn

(3) clarity about salary and working hours

(4) insight into the culture within the company

(5) expectations from the employer

We also formed a testing ground that consists of a variety of organizations: Stichting Alexander, municipalities Gemeente Amsterdam and Gemeente Etten-Leur, universities Universiteit van Amsterdam (UVA) and Rijksuniversitiet Groningen (RUG), and vacancy platform Stadsvacaturebank. Our partners all have specific expertise with which they will help us in developing our prototype. We see the municipalities as official examples and precursors within the field of labor discrimination. Within the RUG we will work together with scientific experts from the social and behavioral sciences faculty that have specialized in labor discrimination. The Civic AI Lab, an initiative of the UVA, focuses on the application of AI in socially relevant topics. Our collaboration with Stadsvacaturebank focuses on applying A/B testing, so that we can quantify the attractive power of our product.

The solution: VICTORY for more Inclusive Vacancies

Our solution is VICTORY (Vacancy Inclusive Check Track Open Response to You): a vacancy scanner that improves inclusion within vacancies by detecting unwanted, discriminating, and excluding language using AI. For each of the aforementioned themes, we will develop an automated detection solution using Natural Language Processing (NLP), using input from our partners within the testing ground. NLP is a sub field of AI concerning interactions between computers and human language. By combining these models, we plan to obtain a prototype for VICTORY. This prototype will then be tested by the affiliated municipalities and Stadsvacaturebank.

We focus on our tool to be easily accessible and user-friendly. After uploading a vacancy, unwanted and/or non-inclusive text is automatically detected. By clicking on the highlighted text, suggestions for appropriate alternatives are given that can directly be used as substitutes for the highlighted text. An example of this is visualized in Figure 1.

Figure 1: A visualization of VICTORY where non-inclusive words (highlighted in red) and unwanted words (highlighted in orange) are automatically recognized and suggestions for substitutes can be given by clicking on a highlighted word.

The impact and benefits for clients: an Inclusive Vacancy Market

By integrating VICTORY within existing vacancy platforms, it becomes directly accessible for a wide range of organisations. For these organisations, the costs of checking a vacancy on inclusivity are only a fraction of normal vacancy costs, providing a low entry barrier. This way we foresee an opportunity to change the current vacancy landscape and thereby help to decrease youth unemployment. In the long term, VICTORY should contribute to the awareness of employers that a more diverse employee base is the solution to the current tight labour market.

All in all, with VICTORY we wish to establish an inclusive vacancy market and consequently contribute to a more inclusive society as a whole.

Writers: Max Dieleman & Mathyn Scheerder

Curious what we can do for your organisation?

Would you like to know more about how, in the future, VICTORY could stimulate inclusion and diversity in your organization? Get in touch with Casper Rutjes at casper@amsterdamdatacollective.com, or check our contactpage.