Developing a concept-level knowledge base for sentiment analysis in Singlish

TL;DR

This paper introduces an automatically constructed Singlish sentiment lexicon that links multiword expressions to emotion labels and polarity, using graph-mining and multi-dimensional scaling on affective knowledge from multiple sources.

Contribution

It presents a novel, automatically built concept-level sentiment knowledge base for Singlish, differing from existing resources that rely on manual labeling or general NLP datasets.

Findings

The lexicon effectively associates expressions with emotions and polarity.

The construction method combines graph-mining, multi-dimensional scaling, and ensemble labeling.

The resource enhances sentiment analysis for Singlish language processing.

Abstract

In this paper, we present Singlish sentiment lexicon, a concept-level knowledge base for sentiment analysis that associates multiword expressions to a set of emotion labels and a polarity value. Unlike many other sentiment analysis resources, this lexicon is not built by manually labeling pieces of knowledge coming from general NLP resources such as WordNet or DBPedia. Instead, it is automatically constructed by applying graph-mining and multi-dimensional scaling techniques on the affective common-sense knowledge collected from three different sources. This knowledge is represented redundantly at three levels: semantic network, matrix, and vector space. Subsequently, the concepts are labeled by emotions and polarity through the ensemble application of spreading activation, neural networks and an emotion categorization model.