Fine-tuning Tree-LSTM for phrase-level sentiment classification on a Polish dependency treebank. Submission to PolEval task 2

TL;DR

This paper presents a fine-tuned Tree-LSTM model tailored for phrase-level sentiment analysis on Polish dependency trees, incorporating custom regularization and sub-word embeddings, evaluated in the PolEval competition.

Contribution

It introduces a novel adaptation of Child-Sum Tree-LSTM for morphologically rich languages with custom regularization and enhanced embeddings.

Findings

Effective sentiment classification on Polish dependency trees

Improved model performance with zoneout regularization

Successful application in PolEval competition

Abstract

We describe a variant of Child-Sum Tree-LSTM deep neural network (Tai et al, 2015) fine-tuned for working with dependency trees and morphologically rich languages using the example of Polish. Fine-tuning included applying a custom regularization technique (zoneout, described by (Krueger et al., 2016), and further adapted for Tree-LSTMs) as well as using pre-trained word embeddings enhanced with sub-word information (Bojanowski et al., 2016). The system was implemented in PyTorch and evaluated on phrase-level sentiment labeling task as part of the PolEval competition.