# Context-Aware Prediction of Derivational Word-forms

## TL;DR

This paper introduces a neural network model for predicting contextually appropriate derivational word-forms from base lemmas, advancing understanding of morphology in natural language processing.

## Contribution

It presents a novel task of context-aware derivational form prediction and an encoder-decoder neural model tailored for this purpose.

## Key findings

- Model generates valid derivations in context
- Less accurate without lexicon constraints
- Demonstrates potential for morphology modeling

## Abstract

Derivational morphology is a fundamental and complex characteristic of language. In this paper we propose the new task of predicting the derivational form of a given base-form lemma that is appropriate for a given context. We present an encoder--decoder style neural network to produce a derived form character-by-character, based on its corresponding character-level representation of the base form and the context. We demonstrate that our model is able to generate valid context-sensitive derivations from known base forms, but is less accurate under a lexicon agnostic setting.