Siamese convolutional networks based on phonetic features for cognate identification

TL;DR

This paper investigates the application of convolutional neural networks using phonetic features for identifying cognates across languages, demonstrating competitive performance compared to traditional string similarity methods.

Contribution

It introduces a ConvNet architecture tailored for cognate detection based on phonetic features, showing its effectiveness across multiple language families.

Findings

ConvNets achieve competitive results with traditional methods

The approach works well across different language families

Phonetic features improve cognate identification accuracy

Abstract

In this paper, we explore the use of convolutional networks (ConvNets) for the purpose of cognate identification. We compare our architecture with binary classifiers based on string similarity measures on different language families. Our experiments show that convolutional networks achieve competitive results across concepts and across language families at the task of cognate identification.