Hamed Hamed A Alshammari

Mr. Hamed Hamed A Alshammari

22nd International Conference on Machine Learning and Applications (ICMLA 2023), Jacksonville, Florida, USA

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
Verified
Journal Article
Open access
PDF file
HAHamed Hamed A Alshammari
Hamed Hamed A Alshammari
AEAhmed ElSayed
Ahmed ElSayed
KEKhaled Elleithy
Khaled Elleithy
The effectiveness of existing AI detectors is notably hampered when processing Arabic texts. This study introduces a novel AI text classifier designed specifically for Arabic, tackling the distinct challenges inherent in processing this language. A particular focus is placed on accurately recognizing human-written texts (HWTs), an area where existing AI detectors have demonstrated significant limitations. To achieve this goal, this paper utilized and fine-tuned two Transformer-based models, AraELECTRA and XLM-R, by training them on two distinct datasets: a large dataset comprising 43,958 examples and a custom dataset with 3078 examples that contain HWT and AI-generated texts (AIGTs) from various sources, including ChatGPT 3.5, ChatGPT-4, and BARD. The proposed architecture is adaptable to any language, but this work evaluates these models’ efficiency in recognizing HWTs versus AIGTs in Arabic as an example of Semitic languages. The performance of the proposed models has been compared against the two prominent existing AI detectors, GPTZero and OpenAI Text Classifier, particularly on the AIRABIC benchmark dataset. The results reveal that the proposed classifiers outperform both GPTZero and OpenAI Text Classifier with 81% accuracy compared to 63% and 50% for GPTZero and OpenAI Text Classifier, respectively. Furthermore, integrating a Dediacritization Layer prior to the classification model demonstrated a significant enhancement in the detection accuracy of both HWTs and AIGTs. This Dediacritization step markedly improved the classification accuracy, elevating it from 81% to as high as 99% and, in some instances, even achieving 100%.

19 July 2024 in Information
Verified
Journal Article
Open access
PDF file
HAHamed Hamed A Alshammari
Hamed Hamed A Alshammari
KEKhaled Elleithy
Khaled Elleithy
Current AI detection systems often struggle to distinguish between Arabic human-written text (HWT) and AI-generated text (AIGT) due to the small marks present above and below the Arabic text called diacritics. This study introduces robust Arabic text detection models using Transformer-based pre-trained models, specifically AraELECTRA, AraBERT, XLM-R, and mBERT. Our primary goal is to detect AIGTs in essays and overcome the challenges posed by the diacritics that usually appear in Arabic religious texts. We created several novel datasets with diacritized and non-diacritized texts comprising up to 9666 HWT and AIGT training examples. We aimed to assess the robustness and effectiveness of the detection models on out-of-domain (OOD) datasets to assess their generalizability. Our detection models trained on diacritized examples achieved up to 98.4% accuracy compared to GPTZero’s 62.7% on the AIRABIC benchmark dataset. Our experiments reveal that, while including diacritics in training enhances the recognition of the diacritized HWTs, duplicating examples with and without diacritics is inefficient despite the high accuracy achieved. Applying a dediacritization filter during evaluation significantly improved model performance, achieving optimal performance compared to both GPTZero and the detection models trained on diacritized examples but evaluated without dediacritization. Although our focus was on Arabic due to its writing challenges, our detector architecture is adaptable to any language.