Prize: 6000 AZN
Get ready for an exciting race.
The objective of this competition is to develop classification models that can automatically route multilingual reviews written by users at the end of their rides on the Apar platform, which provides bicycle and scooter rental services, to the corresponding departments. The texts within the provided dataset are mainly written in a mix of Azerbaijani, Russian, and English.
Your task is to analyze the feedback and, if appropriate, the tag information by utilizing artificial intelligence, machine learning, or other methods, to correctly predict the corresponding department label (label) for each review in order to rapidly identify user issues and optimize bicycle maintenance processes.
Apar is a micro-mobility platform providing bicycle and scooter rental services in Azerbaijan, creating opportunities for users to move around the city more conveniently, actively, and affordably via a mobile application. In addition to providing an alternative solution to daily transportation, Apar supports the development of micro-mobility culture, the promotion of eco-friendly transport, and the formation of a healthier lifestyle. The main goal of the company is to introduce the active mobility culture to a wide audience and increase interest in micro-mobility within the urban ecosystem.
Submissions will be evaluated using the Macro F1-Score metric.
The total prize pool for this competition is 6000 AZN, distributed as follows:
The dataset provided in this competition is the exclusive property of NAIC and the Competition Host (Apar), protected under the laws of the Republic of Azerbaijan on Copyright and Personal Data:
