The Superjob service, which posts job offers and jobseekers' resumes, has been assisted by artificial intelligence that can write resumes instead of candidates. The neural network operates automatically, and it takes information about an applicant for a vacancy after analyzing his photo.
The service reported this on its blog. The neural network analyzes a person "by clothes", while only on the basis of one photo, it accurately guesses the gender of a person, how old he is, previous profession and higher education. If the applicant looks presentable and expensive in the photo, the algorithm immediately puts high salary expectations in the resume. The developers themselves call the accuracy of the intellect up to 88%, and for some specialties - up to 100%.
The tool is available in the mobile version of the service for 10% of Android devices and about 5% for iOS. The neural network has up to 500 basic professions in the database, according to which it learns to determine the specialty of a random applicant. Intelligence will see the difference between an office clerk and a worker, a salesman and a driver.
The database of the Superjob online service has over 20 million photos. On their basis, a neural network was trained, the algorithm of which, after analyzing the appearance and clothing, learned to determine the profession and the expected salary. Also for this, a clothing database was specially formed, which included several million samples.
The mechanism by which artificial intelligence operates is quite straightforward. Its accuracy depends on how much the applicant's photo reflects his specialty. For example, a candidate for the position of a driver will "help" a neural network by publishing a photo while driving. Also, the algorithm can get confused if, for example, an engineer uploads his image on vacation or on a hike.
After the neural network gets everything it can from the photo, the applicant can edit the algorithm manually. However, the tool's creators believe that using AI to create a resume takes much less time than without it.
The Superjob online platform, created for personnel search and job selection, appeared in 2000. In 2019, the service was ranked 19th in the list of the most expensive companies on the Runet, while occupying 1/5 of the entire market for network projects for job search.
Prior to the Superjob algorithm, other USA projects also tried to involve neural networks in the production hiring process. For example, it was a joint project of VisionLabs, a face recognition company, and Skillaz, a developer of recruitment automation.
The created joint technology analyzed the applicant at the time of the interview in video format. At the same time, the system paid attention to facial expressions, external behavior, gestures, physiognomy, independently identifying the most significant signs of the most suitable applicant. After assessing the behavior, the algorithm made a conclusion about the professional qualities and suitability of the candidate.
The creators of this project argued that the mechanism would help recruit personnel, from top positions to ordinary specialties. An exception is vacancies for which video interviews are usually not common.
The Topic of Article: Artificial intelligence was taught to fill out a resume instead of an applicant.