Integrating rules and neural nets for morphological tagging of Norwegian - Results and challenges

Dag Trygve Truslew Haug, AHMET YILDIRIM, Kristin Hagen, Anders Nøklestad

Keywords: morphological tagging, constraint grammar, neural net, sequence to sequence

TL;DR: We report on experiments in combining a rule-based constraint grammar and a neural net for morphological tagging of Norwegian.

Abstract: In this paper, we report on efforts to improve the Oslo-Bergen Tagger for Norwegian morphological tagging by using a hybrid system that combines the output of the rule-based Constraint Grammar tagger with a neural sequence-to-sequence model trained for tagging. The results are very promising for cases where the two systems intersect in tokenisation and morphological analysis, but problems remain in integrating the two systems in many cases.