Unifying Knowledge, Reasoning, and Hierarchy for Classifying Harmful Content

Authors

  • Raden Budiarto Politeknik Siber dan Sandi Negara
  • I Gede Maha Putra Politeknik Siber dan Sandi Negara

DOI:

https://doi.org/10.29207/resti.v10i1.7115

Keywords:

decision tree-of-thought, harmful content classification, hierarchical text classification, knowledge graphs

Abstract

The spread of negative, engagement-driven content online causes significant societal harm, requiring advanced automated moderation tools. However, current classification systems often treat harmful content subtypes as independent, "flat" categories, which hinders their ability to thematically overlap content. This study designed and validated a novel integrated framework to accurately and transparently classify such complex cases. We proposed KG-DToT-HTC, a hybrid framework that synergistically combines three methodologies: a predefined Hierarchical Text Classification (HTC) taxonomy to structure the decision-making process; a domain-specific Knowledge Graph (KG) to provide factual, real-world context; and Decision Tree-of-Thought (DToT) prompting to guide a Large Language Model through an explicit, step-by-step reasoning process. On a real-world dataset of harmful Indonesian news, the proposed framework achieved a state-of-the-art Macro-F1 score of 0.934, representing a nearly 15-percentage point improvement over a zero-shot baseline. Ablation studies confirmed that each component—hierarchy, knowledge, and reasoning—provided a distinct and critical contribution to the final performance. The major conclusion of this study is that a synergistic architecture is essential for the accurate classification of complex harmful content. This work demonstrates a viable path toward "glass-box," interpretable AI moderation systems whose decisions are not only highly accurate but also fully auditable.

Downloads

Download data is not yet available.

References

C. E. Robertson, N. Pröllochs, K. Schwarzenegger, P. Pärnamets, J. J. Van Bavel, and S. Feuerriegel, “Negativity drives online news consumption,” Nat Hum Behav, vol. 7, no. 5, pp. 812–822, Mar. 2023, doi: 10.1038/s41562-023-01538-4.

J. Watson, S. Van Der Linden, M. Watson, and D. Stillwell, “Negative online news articles are shared more to social media,” Sci Rep, vol. 14, no. 1, p. 21592, Sep. 2024, doi: 10.1038/s41598-024-71263-z.

C. W. Jo, M. Wesołowska, and M. Wojcieszak, “Harmful YouTube Video Detection: A Taxonomy of Online Harm and MLLMs as Alternative Annotators,” Nov. 06, 2024, arXiv: arXiv:2411.05854. doi: 10.48550/arXiv.2411.05854.

A. Sheth, V. L. Shalin, and U. Kursuncu, “Defining and detecting toxicity on social media: context and knowledge are key,” Neurocomputing, vol. 490, pp. 312–318, Jun. 2022, doi: 10.1016/j.neucom.2021.11.095.

A. Toktarova et al., “Hate Speech Detection in Social Networks using Machine Learning and Deep Learning Methods,” IJACSA, vol. 14, no. 5, 2023, doi: 10.14569/IJACSA.2023.0140542.

W. M. Johnston and G. C. L. Davey, “The psychological impact of negative TV news bulletins: The catastrophizing of personal worries,” British J of Psychology, vol. 88, no. 1, pp. 85–91, Feb. 1997, doi: 10.1111/j.2044-8295.1997.tb02622.x.

A. Schonfeld, D. McNiel, T. Toyoshima, and R. Binder, “Cyberbullying and Adolescent Suicide,” J Am Acad Psychiatry Law, vol. 51, no. 1, pp. 112–119, Mar. 2023, doi: 10.29158/JAAPL.220078-22.

N. Ayoobi, S. Shahriar, and A. Mukherjee, “Seeing Through AI’s Lens: Enhancing Human Skepticism Towards LLM-Generated Fake News,” in Proceedings of the 35th ACM Conference on Hypertext and Social Media, Poznan Poland: ACM, Sep. 2024, pp. 1–11. doi: 10.1145/3648188.3675136.

S.-Y. Lin, Y.-C. Kung, and F.-Y. Leu, “Predictive intelligence in harmful news identification by BERT-based ensemble learning model with text sentiment analysis,” Information Processing & Management, vol. 59, no. 2, p. 102872, Mar. 2022, doi: 10.1016/j.ipm.2022.102872.

S. Raza, D. Paulen-Patterson, and C. Ding, “Fake news detection: comparative evaluation of BERT-like models and large language models with generative AI-annotated data,” Knowl Inf Syst, vol. 67, no. 4, pp. 3267–3292, Apr. 2025, doi: 10.1007/s10115-024-02321-1.

X. Men and V. Y. Mariano, “Explainable Fake News Detection Based on BERT and SHAP Applied to COVID-19,” IJMECS, vol. 16, no. 1, pp. 11–22, Feb. 2024, doi: 10.5815/ijmecs.2024.01.02.

S. Daud, M. Ullah, A. Rehman, T. Saba, R. Damaševičius, and A. Sattar, “Topic Classification of Online News Articles Using Optimized Machine Learning Models,” Computers, vol. 12, no. 1, p. 16, Jan. 2023, doi: 10.3390/computers12010016.

H. B. Aji and E. B. Setiawan, “Detecting Hoax Content on Social Media Using Bi-LSTM and RNN,” bits, vol. 5, no. 1, Jun. 2023, doi: 10.47065/bits.v5i1.3585.

C. Li, G. Zhan, and Z. Li, “News Text Classification Based on Improved Bi-LSTM-CNN,” in 2018 9th International Conference on Information Technology in Medicine and Education (ITME), Hangzhou: IEEE, Oct. 2018, pp. 890–893. doi: 10.1109/ITME.2018.00199.

J. P. Haumahu, S. D. H. Permana, and Y. Yaddarabullah, “Fake news classification for Indonesian news using Extreme Gradient Boosting (XGBoost),” IOP Conf. Ser.: Mater. Sci. Eng., vol. 1098, no. 5, p. 052081, Mar. 2021, doi: 10.1088/1757-899X/1098/5/052081.

S. González-Silot, A. Montoro-Montarroso, E. M. Cámara, and J. Gómez-Romero, “Enhancing Disinformation Detection with Explainable AI and Named Entity Replacement,” Feb. 07, 2025, arXiv: arXiv:2502.04863. doi: 10.48550/arXiv.2502.04863.

J. Chan and Y. Li, “Unveiling disguised toxicity: A novel pre-processing module for enhanced content moderation,” MethodsX, vol. 12, p. 102668, Jun. 2024, doi: 10.1016/j.mex.2024.102668.

W. Zhao, L. Zhu, M. Wang, X. Zhang, and J. Zhang, “WTL-CNN: a news text classification method of convolutional neural network based on weighted word embedding,” Connection Science, vol. 34, no. 1, pp. 2291–2312, Dec. 2022, doi: 10.1080/09540091.2022.2117274.

J. Song, F. Wang, and Y. Yang, “Peer-Label Assisted Hierarchical Text Classification,” in Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Toronto, Canada: Association for Computational Linguistics, 2023, pp. 3747–3758. doi: 10.18653/v1/2023.acl-long.207.

T. Kuzman and N. Ljubešić, “LLM Teacher-Student Framework for Text Classification With No Manually Annotated Data: A Case Study in IPTC News Topic Classification,” IEEE Access, vol. 13, pp. 35621–35633, 2025, doi: 10.1109/ACCESS.2025.3544814.

A. Alamsyah and Y. Sagama, “Empowering Indonesian internet users: An approach to counter online toxicity and enhance digital well-being,” Intelligent Systems with Applications, vol. 22, p. 200394, Jun. 2024, doi: 10.1016/j.iswa.2024.200394.

Y. Liu et al., “Enhancing Hierarchical Text Classification through Knowledge Graph Integration,” in Findings of the Association for Computational Linguistics: ACL 2023, Toronto, Canada: Association for Computational Linguistics, 2023, pp. 5797–5810. doi: 10.18653/v1/2023.findings-acl.358.

Y. Han, A. Silva, L. Luo, S. Karunasekera, and C. Leckie, “Knowledge Enhanced Multi-modal Fake News Detection,” Aug. 10, 2021, arXiv: arXiv:2108.04418. doi: 10.48550/arXiv.2108.04418.

Y. Cao, H. Li, Y. Yu, and S. R. Javaji, “Capybara at the Financial Misinformation Detection Challenge Task: Chain-of-Thought Enhanced Financial Misinformation Detection,” in Proceedings of the Joint Workshop of the 9th Financial Technology and Natural Language Processing (FinNLP), the 6th Financial Narrative Processing (FNP), and the 1st Workshop on Large Language Models for Finance and Legal (LLMFinLegal), C.-C. Chen, A. Moreno-Sandoval, J. Huang, Q. Xie, S. Ananiadou, and H.-H. Chen, Eds., Abu Dhabi, UAE: Association for Computational Linguistics, Jan. 2025, pp. 321–325. [Online]. Available: https://aclanthology.org/2025.finnlp-1.38/

B. Xie et al., “Multiknowledge and LLM-Inspired Heterogeneous Graph Neural Network for Fake News Detection,” IEEE Trans. Comput. Soc. Syst., vol. 12, no. 2, pp. 682–694, Apr. 2025, doi: 10.1109/TCSS.2024.3488191.

Y. Zhao, J. Zhu, C. Xu, Y. Liu, and X. Li, “Enhancing LLM-based Hatred and Toxicity Detection with Meta-Toxic Knowledge Graph,” in Findings of the Association for Computational Linguistics: ACL 2025, Vienna, Austria: Association for Computational Linguistics, 2025, pp. 24747–24760. doi: 10.18653/v1/2025.findings-acl.1270.

A. P. Muharram and A. Purwarianti, “Enhancing Natural Language Inference Performance with Knowledge Graph for COVID-19 Automated Fact-Checking in Indonesian Language,” J. ICT Res. Appl., vol. 19, no. 1, pp. 27–46, Sep. 2025, doi: 10.5614/itbj.ict.res.appl.2025.19.1.2.

H. Wu et al., “ToTRL: Unlock LLM Tree-of-Thoughts Reasoning Potential through Puzzles Solving,” Dec. 26, 2025, arXiv: arXiv:2505.12717. doi: 10.48550/arXiv.2505.12717.

S. Yao et al., “Tree of Thoughts: Deliberate Problem Solving with Large Language Models,” Dec. 03, 2023, arXiv: arXiv:2305.10601. doi: 10.48550/arXiv.2305.10601.

C. Singh, J. Morris, A. Rush, J. Gao, and Y. Deng, “Tree Prompting: Efficient Task Adaptation without Fine-Tuning,” in Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, Singapore: Association for Computational Linguistics, 2023, pp. 6253–6267. doi: 10.18653/v1/2023.emnlp-main.384.

J. Zhang, Q. Wu, Y. Xu, C. Cao, Z. Du, and K. Psounis, “Efficient Toxic Content Detection by Bootstrapping and Distilling Large Language Models,” AAAI, vol. 38, no. 19, pp. 21779–21787, Mar. 2024, doi: 10.1609/aaai.v38i19.30178.

J. Bai et al., “AutoSchemaKG: Autonomous Knowledge Graph Construction through Dynamic Schema Induction from Web-Scale Corpora,” Aug. 01, 2025, arXiv: arXiv:2505.23628. doi: 10.48550/arXiv.2505.23628.

Downloads

Published

2026-02-25

Issue

Section

Artificial Intelligence

How to Cite

Budiarto, R., & I Gede Maha Putra. (2026). Unifying Knowledge, Reasoning, and Hierarchy for Classifying Harmful Content. Jurnal RESTI (Rekayasa Sistem Dan Teknologi Informasi), 10(1), 133-141. https://doi.org/10.29207/resti.v10i1.7115