Home / Scientific Arbitration Methodology
Scientific Arbitration & Linguistic Proofing

The Scientific Methodology for Arbitrating the AI Readiness Assessment

A standardized comparative reading and a comprehensive scientific arbitration of three scientific models, converging on a single arbitrated model of (100) questions across four axes, with a clear scientific approach and distinguished linguistic proofing.

Prepared by: Dr. Islam Mohammed Youssef Approved First Report — 9 May 2026
Scientific Methodology

Work Methodology

Reviewing the questions linguistically in the field of artificial intelligence, and arbitrating them scientifically through specialization and expertise, is a methodological cornerstone for ensuring the quality of the measurement, evaluation, and applied tools. It safeguards the accuracy of the scientific content and its alignment with learning objectives and higher-order thinking skills. This academic analysis and linguistic proofing provide a standardized comparative reading and a comprehensive scientific arbitration of three scientific models, converging on a single arbitrated model of (100) questions across four axes, with a clear scientific approach and distinguished linguistic proofing.

The Four Arbitration Principles

1
Precise Question Construction

Each question is worded precisely so that it aligns with the target and the measured domain, with a sample that is representative of the knowledge and skill domain.

2
A Clear Scientific Approach to the Options

A single clearly correct alternative, with plausible distractors that are neither far-fetched nor obvious, and avoidance of ambiguity and misleading wording.

3
Knowledge Value in the AI Context

Content aligned with the correct concepts of artificial intelligence and the adopted academic standards, with the question tied to the required knowledge level.

4
Distinguished Linguistic Proofing

Reviewing the scientific material linguistically for hamzas, punctuation, and rephrasing, with sound spelling, grammar, and semantics.

The Four Assessment Axes

Knowledge
25 questions
Skill
25 questions
Awareness
25 questions
Bias & Fairness
25 questions

How It Is Used

  1. Enter the reviewer's data in the designated fields.
  2. Paste the questions, options, and current answer into the "Review" sheet.
  3. The reviewer fills in the evaluation and final-result columns, then adds the proposed edits when needed.
  4. Flag the cells that require an edit, even if the edit text has not yet been written.
  5. Prepare the model for later merging, since each question has a fixed number, a unified result, and supporting columns.
  6. Distinguished linguistic proofing (spelling, grammar, and semantics) and punctuation.

Adopted Best Practices

  • Align the question with the target or measured domain, with a sample representative of the knowledge/skill domain.
  • Ensure the question stem is clear and focused and allows the problem to be understood before reading the alternatives.
  • Ensure a single clearly best alternative, with plausible distractors that are neither far-fetched nor obvious.
  • Avoid ambiguity, misleading wording, and unnecessary negative statements.
  • Check the question's classification within the appropriate axis, and verify the correctness of the current answer.

Most Frequent Errors & Notes

Semantic Errors

Such as: what is meant by, and rephrasing questions and answers.

Grammatical Errors

Such as the correct use of interrogatives ("What is…?").

Spelling Errors

Such as hamzas and punctuation marks.

Key Arbitration & Proofing Actions

  • Adopting a model of (100) questions by rephrasing the questions and reviewing their options.
  • Correcting the linguistic errors (semantic, grammatical, and spelling) in the questions and answers.
  • Unifying the terminology across the questions and answers.
  • Producing a unified final model of (100 questions and 300 reviewed options) ready for application and use.
Arbitration Committee

Reviewers & Their Profiles

The scientific arbitration and linguistic proofing were carried out by a select group of academics specialized in artificial intelligence and the Arabic language. Their profiles are below.

A
Dr. Abdullah Ibrahim Alharbi Scientific arbitration
Arabic NLP Scientist · NVIDIA-Certified Instructor · Researcher
Natural Language Processing & Large Language Models (LLMs) King Salman Global Academy for Arabic Language, Riyadh · University of Birmingham

An expert in NLP and AI research & development, holding a PhD in Computer Science and over 15 years of experience in applied AI and NLP research and engineering. He currently leads national-scale linguistic-data and AI R&D initiatives, delivering large-scale Arabic datasets, LLM-powered systems, and semantic-search platforms. He is an NVIDIA-certified instructor in large language models, RAG systems, and Arabic dialect modeling.

Education
  • PhD in Computer Science — University of Birmingham, UK (2018–2022) · Focus: Arabic NLP and Large Language Models.
  • MSc in Computer Science — The University of Queensland, Australia (2012–2014) · Information Retrieval and Machine Learning.
  • BSc in Computer Science — Qassim University, Saudi Arabia (2001–2006).
Experience
  • Manager, Arabic Language Resource Development — King Salman Global Academy for Arabic Language (2025–Present).
  • Assistant Professor (part-time) — Imam Muhammad ibn Saud Islamic University (2025–Present).
  • Assistant Professor — King Abdulaziz University (2022–2025).
  • Head of Standardized Testing, Partnerships & Cooperative Training Units — King Abdulaziz University (2023–2025).
  • Research Associate — University of Birmingham (2022–2023).
Research Interests
  • Large language models and generative AI, hallucination detection, and semantic understanding.
  • Arabic linguistic data and speech resources (a Saudi-dialect dataset exceeding 2,500 hours), semantic search, and RAG systems.
Professional Certifications
  • NVIDIA-Certified Instructor (2025–2027) in LLMs, Transformers, RAG, and deep learning.
  • NVIDIA University Ambassador (2025–2027).
  • NVIDIA-Certified Associate: Generative AI — LLMs, and Multimodal.
  • IBM-Certified Instructor in Data Science.
Technical Skills
  • Transformers · BERT · GPT · LLaMA · AraBERT · SBERT · RAG systems · semantic search.
  • PyTorch · HuggingFace · Python · FastAPI · SQL/NoSQL · Docker.
Awards
  • Excellence Award (three times) — Saudi Cultural Bureau, UK (2020–2022).
  • Second Place — Arabic Sentiment Analysis Challenge (KAUST 2021).
  • Best Paper Award — NLDB Conference (2020).
H
Dr. Hanan Nasser Muhajab Scientific arbitration
Assistant Professor — College of Engineering & Computer Science, Jazan University
Artificial Intelligence, Knowledge Graphs & the Semantic Web Jazan University, Saudi Arabia

An Assistant Professor at the College of Engineering and Computer Science, and a researcher specialized in artificial intelligence, knowledge graphs, the semantic web, and linked data, with experience in geographic data modeling and data science.

Education
  • Associate Fellow of Advance HE (AFHEA) — Cardiff University, UK (March 2025).
  • PhD in Computer Science — Cardiff University, UK (2021–2025).
  • MSc in Computer Science — Kent State University, USA (2014–2016) · GPA 4.0/4.0.
  • BSc in Computer Science — Taif University, Saudi Arabia (2006–2010) · GPA 3.78/4.0.
Experience
  • Assistant Professor — Jazan University (2025–2026) · Faculty Professional Development Coordinator and member of the Engineering & Technology Research Center.
  • Teaching Assistant — Cardiff University (2021–2025) · supervising research in knowledge graphs and linked data.
  • Research Assistant — Kent State University (2018–2019) · eye-tracking and translation project.
  • Lecturer — Jazan University (2017–Present).
Skills & Interests
  • Knowledge-graph techniques and RDF stores: Neo4j · Stardog · GraphDB · Apache Jena.
  • Graph query languages: SPARQL · GEOSPARQL · Python · machine-learning algorithms · graph embedding.
Selected Publications
  • CyberGraphNet: Integrating Knowledge Graphs and Neural Networks for Cybersecurity Analytics (IEEE WiDS PSU 2026).
  • Unified Management of Place Information on the Web of Data (PhD dissertation, Cardiff University 2025).
  • Spatial Semantics for the Evaluation of Administrative Geospatial Ontologies (ISPRS IJGI 2024).
Talks & Participation
  • Represented Jazan University in an open dialogue hosted by the National Defense College, Riyadh (April 2026): "Scientific Research: Regulations for the Use of Generative AI."
  • Delivered the seminar "Innovation and Invention in the Age of Artificial Intelligence" at the Engineering & Technology Research Center, Jazan University (March 2026).
Awards
  • Certificate of Achievement (January 2026) for contributions to the Engineering & Technology Research Center.
  • Certificate of Appreciation (September 2025) for participating in the Women in Data Science forum.
H
Ms. Hala Abaker Ismail Zain Scientific arbitration
Lecturer — College of Engineering & Computer Science, Jazan University
Information Systems, Cloud Computing, Networks & Data Science Jazan University, Saudi Arabia

A lecturer at the College of Engineering and Computer Science and a PhD candidate in cloud computing and networks, with extensive academic and administrative experience in information systems, graduation-project arbitration, curriculum development, and student activities.

Education
  • PhD Candidate — University of Malaya, Malaysia · Cloud Computing and Networks (since March 2024).
  • MSc in Information Systems — University of Michigan, USA (December 2015) · GPA 3.97/4.
  • BSc in Information Systems — Imam Muhammad ibn Saud Islamic University, Saudi Arabia (February 2010) · GPA 4.10/5.
Experience
  • Assistant to the Head of the IT & Information Security Department, Female Section (April 2021 – May 2023).
  • Graduation-project arbitration committee member, Computer Science and IT & Information Security departments (since August 2016).
  • Head of the Student Activities Unit, Female Section (October 2019 – April 2021).
  • Lecturer at the College of Computer Science and Information Technology (since May 2016).
  • Member of the curriculum-development, examinations, and academic-advising committees.
Workshops & Courses Delivered
  • Summarizing a training package on artificial intelligence · thinking skills · the Six Thinking Hats.
  • Technical report writing · academic research skills · preparing presentations for academic purposes.
Skills
  • Languages: Arabic and English.
  • Computing: Windows/iOS · Office · Visual Basic/Studio · Oracle · C++ · MySQL · JavaScript · PHP · HTML · Python · Hadoop.
  • Decision-making · time management · planning & organization · creative thinking · problem solving.
Publications
  • Cloud Computing Architecture and Services (IOSR-JCE 2017).
  • Framework of Strategic Alignment Through Enterprise Architecture for Organizational Performance (ICSECS 2021).
  • COVID-19 Analysis and Predictions for KSA Using Machine Learning (ICSECS 2021).
  • Deepfakes: Benefits, Threats, and Combat Strategies (book chapter, 2023).
I
Dr. Islam Mohammed Ali Youssef Linguistic proofing · Report author
International Training Consultant · Academic & Educational Supervisor · Linguistic Proofreader
Linguistic & Grammatical Studies, Arabic Language & Academic Supervision Over 24 years of experience in education and training

An international training consultant, academic and educational supervisor, verbal-abilities and national "Nafes" test trainer, Arabic language teacher, head of the identity-subjects department, linguistic proofreader, and school-accreditation consultant. He combines distinguished teaching, organized supervision, and educational research, and is the author of the AI assessment arbitration report and the person responsible for its linguistic proofing.

Education
  • Diploma in Academic Supervision & Research Guidance Skills for Students — Scientific Events Gate, Malaysia · Excellent (2025).
  • Diploma in Leadership in Quality Management — Emalonis Academy, Cairo · Excellent (2022).
  • PhD in Linguistic & Grammatical Studies — Faculty of Arts, Helwan University · Excellent with First-Class Honors (2016).
  • MA in Modern Linguistics & Grammar — Faculty of Arts, Alexandria University · Excellent (2011).
  • General Educational Diploma in Education — Faculty of Education, Alexandria University (2003).
  • BA in Arabic Language & Islamic Studies — Dar Al-Ulum, Cairo University (1999).
Professional License
  • Educational Professional License from the Education & Training Evaluation Commission — "Practitioner Teacher" level · Arabic Language specialization (valid until 31/12/2028).
Professional Roles
  • International training consultant · academic and educational supervisor.
  • Verbal-abilities and national "Nafes" test skills trainer.
  • Head of the identity-subjects department · linguistic proofreader · school-accreditation consultant.
Professional Goal & Values
  • Excellence in academic and educational supervision aligned with 21st-century skills, using technology and AI in development.
  • His values: mastery & development · precision & excellence · creativity & distinction · respect & appreciation · technology & digital transformation.
Results

First Report Results

The key statistics of the scientific arbitration and linguistic proofing across the three models, and the final approved model.

300
questions arbitrated & proofread
across three scientific models
29
questions accepted by all
as-is, without changes
71
questions modified
in wording or options
100
questions finally approved
the model approved scientifically & linguistically

Arbitration & Approval Timeline

Reviewer / Proofreader Role Received On Status
Dr. Abdullah Ibrahim Alharbi Scientific arbitration 28 / 4 / 2026 Approved
Dr. Hanan Nasser Muhajab Scientific arbitration 26 / 4 / 2026 Approved
Ms. Hala Abaker Zain Scientific arbitration 26 / 4 / 2026 Approved
Dr. Islam Mohammed Youssef Linguistic proofing 9 / 5 / 2026 Approved

Axis Distribution in the Approved Model

Knowledge
25 / 100
Skill
25 / 100
Awareness
25 / 100
Bias & Fairness
25 / 100

The Final Approved Version

The scientific arbitration and linguistic proofing were completed to a high standard, and the final model was approved on Saturday, 22 Dhu al-Qi'dah 1447 AH, corresponding to 9 May 2026. It consists of 100 questions with a total of 300 reviewed options (multiple choice), ready for scientific accreditation and application.

A clearer picture. A confident next step.

Understand your AI readiness and find the right path forward.

Explore assessments