Sentiment Classification of Food Reviews

TL;DR

This paper explores neural network approaches for sentiment analysis of food reviews, focusing on model tuning, handling skewed data, and evaluating accuracy improvements.

Contribution

It introduces two methods for managing skewed review data and compares RNN and GRU models for improved sentiment classification.

Findings

GRU outperforms simple RNN in accuracy

Proposed skewed data handling methods improve model performance

Models achieve high accuracy on food review sentiment classification

Abstract

Sentiment analysis of reviews is a popular task in natural language processing. In this work, the goal is to predict the score of food reviews on a scale of 1 to 5 with two recurrent neural networks that are carefully tuned. As for baseline, we train a simple RNN for classification. Then we extend the baseline to GRU. In addition, we present two different methods to deal with highly skewed data, which is a common problem for reviews. Models are evaluated using accuracies.