MyProfessors: Mining Turkish Student Reviews

TL;DR

This paper introduces Hocalarim, a large Turkish student review dataset, analyzes its properties, and explores how institution type and student bias influence ratings.

Contribution

It provides the first extensive Turkish student review dataset and analyzes factors affecting review ratings, such as institution type and bias.

Findings

Institution type impacts review ratings.

Student bias correlates with positive or negative feedback.

Dataset contains over 5000 reviews for research use.

Abstract

We introduce Hocalarim (MyProfessors), the largest student review dataset available for the Turkish language. It consists of over 5000 professor reviews left online by students, with different aspects of education rated on a scale of 1 to 5 stars. We investigate the properties of the dataset and present its statistics. We examine the impact of students' institution type on their ratings and the correlation of students' bias to give positive or negative feedback.