Automatically generating question-answer pairs for assessing basic reading comprehension in Swedish

TL;DR

This paper evaluates a lightweight, data-driven method called Quinductor for automatically generating reading comprehension questions in Swedish, demonstrating its effectiveness as a baseline for neural approaches.

Contribution

It introduces Quinductor, a non-neural, data-driven question generation method for Swedish, and evaluates its quality as a baseline for neural QG models.

Findings

Quinductor is a viable question generation method.

It provides a strong baseline for neural QG methods.

The evaluation confirms its effectiveness in Swedish reading comprehension.

Abstract

This paper presents an evaluation of the quality of automatically generated reading comprehension questions from Swedish text, using the Quinductor method. This method is a light-weight, data-driven but non-neural method for automatic question generation (QG). The evaluation shows that Quinductor is a viable QG method that can provide a strong baseline for neural-network-based QG methods.