Fork me on GitHub

Character-based recurrent neural networks for morphological relational reasoning

Character-based recurrent neural networks for morphological relational reasoning. The <em>FC relation</em> layer is connected to an auxilliary output layer, trained to predict a label for the current type of relation. The final output is generated by the <em>Decoder RNN</em>.

We present a model for predicting word forms based on morphological relational reasoning with analogies. While previous work has explored tasks such as morphological inflection and reinflection, these models rely on an explicit enumeration of morphological features, which may not be available in all cases. To address the task of predicting a word form given a demo relation (a pair of word forms) and a query word, we devise a character-based recurrent neural network architecture using three separate encoders and a decoder. We also investigate a multiclass learning setup, where the prediction of the relation type label is used as an auxiliary task. Our results show that the exact form can be predicted for English with an accuracy of 94.7%. For Swedish, which has a more complex morphology with more inflectional patterns for nouns and verbs, the accuracy is 89.3%. We also show that using the auxiliary task of learning the relation type speeds up convergence and improves the prediction accuracy for the word generation task.

Source code

The source code used for the experiments can be downloaded from

Olof Mogren, Richard Johansson

To appear at Subword & Character Level Models in NLP (SCLeM) workshop at EMNLP 2017 in Copenhagen, Denmark, September 7.
PDF Fulltext bibtex.

Olof Mogren Decisions, networks and analytics lab (DNA), RISE SICS

LinkedIn Twitter Atom/RSS Feed