An Uncertainty-Aware Hybrid Symbolic-Lexicon Framework for Multi-Domain Sentiment Classification
Abstract
Multi-domain sentiment classifiers may return a polarity label even when the evidence behind that decision becomes fragile under domain shift. This study evaluates an uncertainty-aware hybrid symbolic-lexicon framework designed to pair each polarity prediction with a traceable accept-or-defer status. The framework combines a lexical branch adapted only on training data with training-only fitted symbolic feature states, variable-precision rough evidence, and nested score fusion. Evaluation uses the UCI Sentiment Labelled Sentences dataset, comprising 3,000 balanced English sentences from Amazon Cells, IMDb, and Yelp, under repeated within-source, pooled multi-domain, and leave-one-source-out protocols. In pooled evaluation, the framework achieved 83.73% accuracy and 83.76% F1-score, improving F1 by 0.50 points over the adaptive lexical branch. The boundary mechanism deferred 53.55% of cases and retained 92.54% accuracy among accepted predictions; however, matched-coverage logistic regression and SVM confidence rankings produced lower selective risk. The contribution is therefore a reproducible, algorithmically traceable polarity-and-deferment mechanism rather than a claim of universal predictive or uncertainty-ranking superiority.
Related
- A multi-stage framework for emotionally intelligent dialogue systems using topological sentiment encoding and causal policy optimization
- Digital Transformer for Fracture Classification: SYM-Fractron, a Physics-Driven Hybrid Binary–Symbolic Computing Ecosystem
- A Symbolic Differentiation_Summation Framework for Power Sums of Arithmetic Sequences
- Adaptive Conformal Redistribution for Inter-class Transitional Uncertainty in Medical Image Classification
- llmClassificR: Text Classification with Large Language Models, Network Psychometrics, and Calibrated Uncertainty in R
BibTeX
@article{yaakub2026uncertaintyaware,
title = {{An Uncertainty-Aware Hybrid Symbolic-Lexicon Framework for Multi-Domain Sentiment Classification}},
author = {Saleh Yaakub and Mustafa Mat Deris and Ika Safitri Windiarti},
journal = {Journal of Information Systems and Informatics},
year = {2026},
month = oct,
doi = {10.63158/journalisi.v8i5.1872},
url = {https://doi.org/10.63158/journalisi.v8i5.1872},
note = {CC BY},
}
RIS
TY - JOUR
TI - An Uncertainty-Aware Hybrid Symbolic-Lexicon Framework for Multi-Domain Sentiment Classification
AU - Saleh Yaakub
AU - Mustafa Mat Deris
AU - Ika Safitri Windiarti
PY - 2026
DA - 2026/10/24
JO - Journal of Information Systems and Informatics
DO - 10.63158/journalisi.v8i5.1872
UR - https://doi.org/10.63158/journalisi.v8i5.1872
AB - Multi-domain sentiment classifiers may return a polarity label even when the evidence behind that decision becomes fragile under domain shift. This study evaluates an uncertainty-aware hybrid symbolic-lexicon framework designed to pair each polarity prediction with a traceable accept-or-defer status. The framework combines a lexical branch adapted only on training data with training-only fitted symbolic feature states, variable-precision rough evidence, and nested score fusion. Evaluation uses the UCI Sentiment Labelled Sentences dataset, comprising 3,000 balanced English sentences from Amazon Cells, IMDb, and Yelp, under repeated within-source, pooled multi-domain, and leave-one-source-out protocols. In pooled evaluation, the framework achieved 83.73% accuracy and 83.76% F1-score, improving F1 by 0.50 points over the adaptive lexical branch. The boundary mechanism deferred 53.55% of cases and retained 92.54% accuracy among accepted predictions; however, matched-coverage logistic regression and SVM confidence rankings produced lower selective risk. The contribution is therefore a reproducible, algorithmically traceable polarity-and-deferment mechanism rather than a claim of universal predictive or uncertainty-ranking superiority.
SN - 2656-5935
SN - 2656-4882
ER -