Poetry generation for Bahasa Indonesia using a constraint satisfaction approach

This paper describes work on a poetry generator that is capable of generating poems in Indonesian based on certain contexts by employing a constraint satisfaction approach. The system retrieves language resources such as templates and slot fillers and combines them to instantiate poetry lines, which in turn are composed into poems based on a set of given constraints. The output of this system was evaluated through an online questionnaire involving 180 respondents. The results showed that poems generated using the full set of constraints were consistently measured better on all aspects than those generated using loose or no constraints.