The Software Department at the College of Information Technology Discusses a Ph.D. Dissertation on "Enrichment of Semantic Annotation Based on Linked Open Data"

Duhaa Fadill Abbas
As part of the ongoing academic activities at the Software Department, College of Information Technology – University of Babylon, a Ph.D. dissertation titled "Enrichment of Semantic Annotation Based on Linked Open Data" was discussed by the doctoral candidate Khalid Jawad Kadhim Abd, under the supervision of Prof. Dr. Asaad Sabah Hadi, at 9:00 AM on Thursday, July 3rd, 2025, in the college’s main conference hall.

The dissertation aims to develop an intelligent system for automatic semantic annotation of entities extracted from unstructured text, addressing one of the key challenges in the field of Natural Language Processing (NLP). The study highlights the limitations of traditional manual annotation approaches, which are time-consuming and often suffer from inconsistencies due to subjective interpretations among annotators.

To overcome these challenges, the researcher proposed an integrated framework that combines Named Entity Recognition (NER), Relation Extraction (RE), Ontology construction, and entity enrichment using Linked Open Data (LOD). The system leverages SPARQL queries to enhance semantic linking accuracy and mitigate ambiguity caused by polysemy and contextual variation.

Evaluation results demonstrated that the system achieved a coverage rate of 79.5% in linking entities to reliable external data sources. Furthermore, it attained an average semantic similarity score of 93% using the BERT model, indicating strong alignment between the extracted context and the enriched data from LOD. Experimental outcomes also confirmed the error-free construction of the ontology, with accurately defined relationships and a total reasoning time of 2.49 seconds.

The dissertation concludes that the proposed system facilitates the construction of RDF triples compatible with semantic web data, thereby enhancing contextual understanding and knowledge representation. For future work, the researcher recommends the incorporation of multiple ontologies instead of relying on a single one, to better address complex relationships among hierarchical concepts. Additionally, a dynamic semantic annotation system is proposed, capable of retrieving real-time explanatory data from various online sources to provide rich and interactive information delivery.