Sentiment Polarity Classification for Khmer

Sentiment polarity classification is a natural language processing (NLP) technique that determines whether a given piece of text has a positive, negative, or neutral sentiment. This article investigates the application of a well-known shallow neural network classifier, FastText, to the Khmer languag...

Full description

Saved in:
Bibliographic Details
Main Authors: Khim, Sokheng, Thu, Ye Kyaw, Sam, Sethserey
Format: Conference Proceeding
Language:English
Subjects:
Benchmark testing
FastText
low-resource NLP
Machine learning
Neural networks
polarity classification
Sentiment analysis
Support vector machines
traditional machine learning
Training
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Sentiment polarity classification is a natural language processing (NLP) technique that determines whether a given piece of text has a positive, negative, or neutral sentiment. This article investigates the application of a well-known shallow neural network classifier, FastText, to the Khmer language, which is a low-resource language. To our knowledge, there isn't a freely available polarity corpus for Khmer. Thus, one of the primary objectives of this research is to develop a human-annotated benchmark corpus tailored for sentiment analysis in this language. Our findings indicate that the FastText model achieves a precision and recall of 0.740 using our modest-sized Khmer polarity corpus, which consists of 9,000 training sentences and 1,000 testing sentences. Among traditional machine learning methods, the support vector machine (SVM) outperforms the rest.
ISSN:2831-4565
DOI:10.1109/iSAI-NLP60301.2023.10354988