PSUT Portal Project Title: Towards Intelligent Legal Systems: Integrating Knowledge Graphs and AI for Arabic Legal Text Understanding
Responsible Researcher: Dr. Omar Alqawasmeh
Project Description:
Legal texts in the Arabic language pose significant challenges for computational analysis due to their complexity, ambiguity, and domain-specific terminology. The highly formal structure, frequent use of archaic and context-dependent expressions, and extensive cross-referencing of laws make Arabic legal language particularly resistant to shallow text processing methods. Traditional Natural Language Processing (NLP) techniques often fail to capture the semantic depth, contextual dependencies, and logical reasoning required for tasks such as legal document classification, precedent retrieval, and contract analysis.
In addition, the scarcity of large-scale, high-quality, and annotated Arabic legal corpora significantly limits the performance of machine learning models. While English and European legal domains benefit from ontologies, structured datasets, and well-established digital legal repositories, Arabic legal systems remain underrepresented in computational law research. This gap not only hinders the development of intelligent legal information systems in Arabic-speaking countries but also creates barriers to legal accessibility, transparency, and innovation. Addressing this issue requires a hybrid approach that leverages both AI-driven NLP models (to process and extract meaning from unstructured texts) and Knowledge Graphs (to organize legal entities, relationships, and reasoning structures into machine-readable semantic networks). Such integration has the potential to transform Arabic legal systems into more transparent, accessible, and intelligent infrastructures.
The overarching research problem can therefore be summarized as:
How can the integration of AI-driven NLP models and Knowledge Graphs enhance the semantic representation, reasoning, and accessibility of Arabic legal texts?
Detailed research questions:
- How can automated entity recognition and relation extraction methods be adapted to accurately capture the unique linguistic and structural features of Arabic legal texts?
- What reasoning and inference mechanisms, supported by Knowledge Graphs, can best assist in tasks such as legal precedent retrieval, contract analysis, and compliance checking in Arabic law?
- To what extent can the proposed system improve legal information accessibility and usability for different stakeholders (judges, lawyers, policymakers, and the public) compared to existing baseline systems?
Application Requirements:
The applicant must have a very good level of mathematics and strong Python programming skills, in addition to meeting the Emerging Researchers Scholarship application requirements.
Application Procedure:
Interested applicants should submit their CV along with all required documents to the following email address:
graduateschool@psut.edu.jo
For more information about the program, please visit: Emerging Researcher Scholarship
Important Dates:
Deadline for Master’s applications: 20 September 2025
Start of the First Semester 2025/2026: 5 October 2025