Integrated ensemble of BERT- and features-based models for authorship attribution in Japanese literary works

2025-04-11

AI Summary

This paper addresses the few-shot Japanese literary author attribution (AA) task. Methodologically, it proposes a dual-path ensemble framework integrating traditional stylistic features with pre-trained language models (PLMs). It is the first to empirically validate BERT’s effectiveness for few-shot Japanese AA; simultaneously, TF-IDF and character n-gram features are extracted and modeled using multi-layer classifiers—XGBoost, SVM, and MLP—with weighted voting for fusion. The core contribution lies in establishing a synergistic “PLM + traditional features” dual-path paradigm, substantially improving generalization under data-scarce conditions. Experiments on test sets disjoint from pre-training data demonstrate an approximately 14-percentage-point improvement in macro-F1 score over baseline models. The integrated model consistently outperforms all individual components, establishing a new state-of-the-art performance ceiling for few-shot Japanese AA.

Technology Category

• Natural Language Processing: Sentiment Analysis, Stylistic Analysis, and Argument Mining
• Machine Learning: Ensemble Methods
• Cognitive Modeling & Cognitive Systems: Computational Creativity

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Traditionally, authorship attribution (AA) tasks relied on statistical data analysis and classification based on stylistic features extracted from texts. In recent years, pre-trained language models (PLMs) have attracted significant attention in text classification tasks. However, although they demonstrate excellent performance on large-scale short-text datasets, their effectiveness remains under-explored for small samples, particularly in AA tasks. Additionally, a key challenge is how to effectively leverage PLMs in conjunction with traditional feature-based methods to advance AA research. In this study, we aimed to significantly improve performance using an integrated integrative ensemble of traditional feature-based and modern PLM-based methods on an AA task in a small sample. For the experiment, we used two corpora of literary works to classify 10 authors each. The results indicate that BERT is effective, even for small-sample AA tasks. Both BERT-based and classifier ensembles outperformed their respective stand-alone models, and the integrated ensemble approach further improved the scores significantly. For the corpus that was not included in the pre-training data, the integrated ensemble improved the F1 score by approximately 14 points, compared to the best-performing single model. Our methodology provides a viable solution for the efficient use of the ever-expanding array of data processing tools in the foreseeable future.

Problem

• Improving authorship attribution in small Japanese literary samples
• Combining BERT and traditional feature-based methods effectively
• Enhancing performance via integrated ensemble models for AA tasks

Innovation

• Integrated ensemble of BERT and feature-based models
• Combines traditional and modern text classification methods
• Improves F1 score significantly in small-sample tasks