Table of Contents
Open Table of Contents
TL; DR
This research project was the second project I conducted with the PEEC research group. In addition to my full-time job at Allplan Software Engineering GmbH, I also worked as a software developer for 10 hours per week, exploring various ways to enhance and find new ways to improve the recruiting system of karriere.at. Our primary focus was on improving the recommendation system, which suggests jobs to users based on their interests. To achieve this, we investigated various approaches for generating job suggestions that take into account dynamically changing interests of a user.
How I came to the project?
Shortly before the completion of the previous project, I was asked by the research group whether I would be interested in a further project. Again, the task immediately convinced me, which is why I wanted to participate in this project as well. Everyone in Austria is familiar with karriere.at and I had already used it, so I was very interested in the topic. However, in the meantime I was again employed full-time at Allplan, which is why I decided to work on the project for only 10 hours per week. Here, too, I was working as a software developer.
The project
The task in the project was clearly defined. The goal of this project was to investigate alternative or new possibilities to improve the recruiting system of karriere.at. The focus was on the area of the recommendation system which suggests suitable jobs for users. New and alternative approaches are to be investigated in order to improve the suggestions of jobs for users.
The task is particularly exciting because it addresses a very general problem. Many people know that the search for a new job can be very time consuming. Therefore, it would be an advantage if platforms like karriere.at could already suggest the perfect job without time-consuming searches. However the perfect job should refer exactly to the user. But therefore, the user should not endlessly type in information to get exactly the right suggestion. The system should independently determine or learn which jobs the job seeker is interested in.
In order to respond to the changing interests of the user, a growing candidate profile has been designed. This profile includes the user’s interests over time, for example.
It is important to emphasize that the consideration of user privacy played a crucial role in the design of this user profile.
After the user profile has been created, it is used to provide the user with job suggestions that match their interests.
To implement the prototype, we primarily utilized the Python programming language and had the opportunity to experiment with various frameworks and services. Furthermore, we incorporated multiple data sources and databases into this project.
By the end of the project, we were able to present several options to enhance the existing Recommendation System project.