LinguaSynth: Heterogeneous Linguistic Signals for News Classification

TL;DR

LinguaSynth introduces a transparent, resource-efficient text classification framework that integrates multiple linguistic signals, achieving high accuracy and interpretability without relying on complex neural networks.

Contribution

It presents a novel framework combining five linguistic feature types within a logistic regression model, enhancing interpretability and efficiency in NLP tasks.

Findings

Achieves 84.89% accuracy on 20 Newsgroups dataset.

Surpasses TF-IDF baseline by 3.32%.

Identifies the importance of syntactic and entity signals.

Abstract

Deep learning has significantly advanced NLP, but its reliance on large black-box models introduces critical interpretability and computational efficiency concerns. This paper proposes LinguaSynth, a novel text classification framework that strategically integrates five complementary linguistic feature types: lexical, syntactic, entity-level, word-level semantics, and document-level semantics within a transparent logistic regression model. Unlike transformer-based architectures, LinguaSynth maintains interpretability and computational efficiency, achieving an accuracy of 84.89 percent on the 20 Newsgroups dataset and surpassing a robust TF-IDF baseline by 3.32 percent. Through rigorous feature interaction analysis, we show that syntactic and entity-level signals provide essential disambiguation and effectively complement distributional semantics. LinguaSynth sets a new benchmark for interpretable, resource-efficient NLP models and challenges the prevailing assumption that deep neural networks are necessary for high-performing text classification.