Detection of Biased Phrases in the Wiki Neutrality Corpus for Fairer Digital Content Management Using Artificial Intelligence
Detecting biased language in large-scale corpora, such as the Wiki Neutrality Corpus, is essential for promoting neutrality in digital content. This study systematically evaluates a range of machine learning (ML) and deep learning (DL) models for the detection of biased and pre-conditioned phrases....
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-07-01
|
| Series: | Big Data and Cognitive Computing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2504-2289/9/7/190 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849303557983961088 |
|---|---|
| author | Abdullah Muhammad Ateeb Ather Olga Kolesnikova Grigori Sidorov |
| author_facet | Abdullah Muhammad Ateeb Ather Olga Kolesnikova Grigori Sidorov |
| author_sort | Abdullah |
| collection | DOAJ |
| description | Detecting biased language in large-scale corpora, such as the Wiki Neutrality Corpus, is essential for promoting neutrality in digital content. This study systematically evaluates a range of machine learning (ML) and deep learning (DL) models for the detection of biased and pre-conditioned phrases. Conventional classifiers, including Extreme Gradient Boosting (XGBoost), Light Gradient-Boosting Machine (LightGBM), and Categorical Boosting (CatBoost), are compared with advanced neural architectures such as Bidirectional Encoder Representations from Transformers (BERT), Long Short-Term Memory (LSTM) networks, and Generative Adversarial Networks (GANs). A novel hybrid architecture is proposed, integrating DistilBERT, LSTM, and GANs within a unified framework. Extensive experimentation with intermediate variants DistilBERT + LSTM (without GAN) and DistilBERT + GAN (without LSTM) demonstrates that the fully integrated model consistently outperforms all alternatives. The proposed hybrid model achieves a cross-validation accuracy of 99.00%, significantly surpassing traditional baselines such as XGBoost (96.73%) and LightGBM (96.83%). It also exhibits superior stability, statistical significance (paired <i>t</i>-tests), and favorable trade-offs between performance and computational efficiency. The results underscore the potential of hybrid deep learning models for capturing subtle linguistic bias and advancing more objective and reliable automated content moderation systems. |
| format | Article |
| id | doaj-art-4ab7ba6177aa47b08d1a1497c5cd4a2d |
| institution | Kabale University |
| issn | 2504-2289 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Big Data and Cognitive Computing |
| spelling | doaj-art-4ab7ba6177aa47b08d1a1497c5cd4a2d2025-08-20T03:58:25ZengMDPI AGBig Data and Cognitive Computing2504-22892025-07-019719010.3390/bdcc9070190Detection of Biased Phrases in the Wiki Neutrality Corpus for Fairer Digital Content Management Using Artificial IntelligenceAbdullah0Muhammad Ateeb Ather1Olga Kolesnikova2Grigori Sidorov3Centro de Investigación en Computación, Instituto Politécnico Nacional, Mexico City 07738, MexicoDepartment of Computer Science, Bahria University Lahore Campus, Lahore 54600, PakistanCentro de Investigación en Computación, Instituto Politécnico Nacional, Mexico City 07738, MexicoCentro de Investigación en Computación, Instituto Politécnico Nacional, Mexico City 07738, MexicoDetecting biased language in large-scale corpora, such as the Wiki Neutrality Corpus, is essential for promoting neutrality in digital content. This study systematically evaluates a range of machine learning (ML) and deep learning (DL) models for the detection of biased and pre-conditioned phrases. Conventional classifiers, including Extreme Gradient Boosting (XGBoost), Light Gradient-Boosting Machine (LightGBM), and Categorical Boosting (CatBoost), are compared with advanced neural architectures such as Bidirectional Encoder Representations from Transformers (BERT), Long Short-Term Memory (LSTM) networks, and Generative Adversarial Networks (GANs). A novel hybrid architecture is proposed, integrating DistilBERT, LSTM, and GANs within a unified framework. Extensive experimentation with intermediate variants DistilBERT + LSTM (without GAN) and DistilBERT + GAN (without LSTM) demonstrates that the fully integrated model consistently outperforms all alternatives. The proposed hybrid model achieves a cross-validation accuracy of 99.00%, significantly surpassing traditional baselines such as XGBoost (96.73%) and LightGBM (96.83%). It also exhibits superior stability, statistical significance (paired <i>t</i>-tests), and favorable trade-offs between performance and computational efficiency. The results underscore the potential of hybrid deep learning models for capturing subtle linguistic bias and advancing more objective and reliable automated content moderation systems.https://www.mdpi.com/2504-2289/9/7/190bias detectionmachine learningdeep learningtext analysisneutralityBERT |
| spellingShingle | Abdullah Muhammad Ateeb Ather Olga Kolesnikova Grigori Sidorov Detection of Biased Phrases in the Wiki Neutrality Corpus for Fairer Digital Content Management Using Artificial Intelligence Big Data and Cognitive Computing bias detection machine learning deep learning text analysis neutrality BERT |
| title | Detection of Biased Phrases in the Wiki Neutrality Corpus for Fairer Digital Content Management Using Artificial Intelligence |
| title_full | Detection of Biased Phrases in the Wiki Neutrality Corpus for Fairer Digital Content Management Using Artificial Intelligence |
| title_fullStr | Detection of Biased Phrases in the Wiki Neutrality Corpus for Fairer Digital Content Management Using Artificial Intelligence |
| title_full_unstemmed | Detection of Biased Phrases in the Wiki Neutrality Corpus for Fairer Digital Content Management Using Artificial Intelligence |
| title_short | Detection of Biased Phrases in the Wiki Neutrality Corpus for Fairer Digital Content Management Using Artificial Intelligence |
| title_sort | detection of biased phrases in the wiki neutrality corpus for fairer digital content management using artificial intelligence |
| topic | bias detection machine learning deep learning text analysis neutrality BERT |
| url | https://www.mdpi.com/2504-2289/9/7/190 |
| work_keys_str_mv | AT abdullah detectionofbiasedphrasesinthewikineutralitycorpusforfairerdigitalcontentmanagementusingartificialintelligence AT muhammadateebather detectionofbiasedphrasesinthewikineutralitycorpusforfairerdigitalcontentmanagementusingartificialintelligence AT olgakolesnikova detectionofbiasedphrasesinthewikineutralitycorpusforfairerdigitalcontentmanagementusingartificialintelligence AT grigorisidorov detectionofbiasedphrasesinthewikineutralitycorpusforfairerdigitalcontentmanagementusingartificialintelligence |