Haji Gul

Knowledge Graph Completion (KGC) seeks to predict missing entities (e.g., heads or tails) or relationships in knowledge graphs (KGs), which often contain incomplete data. Traditional embedding-based methods, such as TransE and ComplEx, have improved tail entity prediction but struggle to generalize to unseen entities during testing. Textual-based models mitigate this issue by leveraging additional semantic context; however, their reliance on negative triplet sampling introduces high computational overhead, semantic inconsistencies, and data imbalance. Recent approaches, like KG-BERT, show promise but depend heavily on entity descriptions, which are often unavailable in KGs. Critically, existing methods overlook valuable structural information in the KG related to the entities and relationships. To address these challenges, we propose Multi-Context-Aware Knowledge Graph Completion (MuCo-KGC), a novel model that utilizes contextual information from linked entities and relations within the graph to predict tail entities. MuCo-KGC eliminates the need for entity descriptions and negative triplet sampling, significantly reducing computational complexity while enhancing performance. Our experiments on standard datasets, including FB15k-237, WN18RR, CoDEx-S, and CoDEx-M, demonstrate that MuCo-KGC outperforms state-of-the-art methods on three datasets. Notably, MuCo-KGC improves MRR on WN18RR, CoDEx-S, and CoDEx-M datasets by 1.63%, 3.77%, and 20.15% respectively, demonstrating its effectiveness for KGC tasks.

Figure 1 — Overview of the MuCo-KGC model pipeline for predicting the tail entity, given a head entity h and relationship r. The left box shows the head context H_c (union of relations R(h) and neighbouring entities E(h)); the right box shows the relationship context R_c. These contextual features, alongside h and r, are fed into a BERT model with a linear classifier and softmax to generate tail-entity probabilities.

@inproceedings{gul-2025-mucokgc,
    title = "MuCo-KGC: Multi-Context-Aware Knowledge Graph Completion",
    author = "Gul, Haji and Naim, Abdul Ghani and Bhat, Ajaz Ahmad",
    booktitle = "Advances in Knowledge Discovery and Data Mining (PAKDD)",
    year = "2025",
    publisher = "Springer",
    doi = "10.1007/978-981-96-8298-0_1",
    url = "https://dl.acm.org/doi/10.1007/978-981-96-8298-0_1"
}

Accurate prediction of drug-target interactions is critical for accelerating drug discovery. In this work, we frame drug-target prediction as a link prediction task on heterogeneous biomedical knowledge graphs (KGs) that integrate drugs, proteins, diseases, pathways, and other relevant entities. Conventional KG embedding methods such as TransE and ComplEx are hindered by their reliance on computationally intensive negative sampling and their limited generalization to unseen drug-target pairs. To address these challenges, we propose Multi-Context-Aware Sampling (MuCoS), a novel framework that prioritizes high-density neighbours to capture salient structural patterns and integrates these with contextual embeddings derived from BERT. By unifying structural and textual modalities and selectively sampling highly informative patterns, MuCoS eliminates the need for negative sampling, significantly reduces computational overhead, and improves generalization to unseen drug-target pairs and targets. Extensive experiments on the KEGG50k and PharmKG-8k datasets demonstrate that MuCoS outperforms competitive baselines, achieving up to 13% improvement in MRR for general relation prediction on KEGG50k, up to 22% improvement in MRR on PharmKG-8k, and up to 6% improvement in dedicated drug-target relation prediction on KEGG50k.

Figure 1 — Overview of the MuCoS model pipeline for predicting general and drug-target relations and tail entities. The input sequence to BERT combines head (h), head context (H_c), tail (t), tail context (T_c), relation (r), and relation context (R_c), passed through BERT with a linear classifier and softmax to produce probabilities for relations and tails.

Figure 2 — Head context (H_c) construction with sampling. The one-hop head context consists of the relation set R(h) and neighbouring tail entities E(h). Only the top-n tail entities are selected and concatenated based on their density ρ(e) to form the optimized head context.

Figure 3 — Relation context (R_c) construction with sampling. The top-k entities are selected by density ρ, retaining the most informative head-tail pairs, which are then aggregated by concatenation to form the optimized relation context.

@inproceedings{gul-etal-2025-mucos,
    title = "{M}u{C}o{S}: Efficient Drug{--}Target Discovery via Multi-Context-Aware Sampling in Knowledge Graphs",
    author = "Gul, Haji and Naim, Abdul Ghani and Bhat, Ajaz Ahmad",
    booktitle = "Proceedings of the 24th Workshop on Biomedical Language Processing",
    month = aug,
    year = "2025",
    address = "Vienna, Austria",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2025.bionlp-1.27/",
    doi = "10.18653/v1/2025.bionlp-1.27",
    pages = "319--327",
    ISBN = "979-8-89176-275-6"
}

[ DUMMY — replace with the real abstract for this paper. ]

[ DUMMY caption — upload your figure as KG-EDAS.png to Dr_Ajaz_and_Haji_work, or change the name here. ]

[ DUMMY BibTeX — paste the real citation here. ]

[ DUMMY — replace with the real abstract for this paper. ]

[ DUMMY caption — upload your figure as KG-Complexity.png to Dr_Ajaz_and_Haji_work, or change the name here. ]

[ DUMMY BibTeX — paste the real citation here. ]

[ DUMMY — replace with the real abstract for this paper. ]

[ DUMMY caption — upload your figure as When-Metrics.png to Dr_Ajaz_and_Haji_work, or change the name here. ]

[ DUMMY BibTeX — paste the real citation here. ]

[ DUMMY — replace with the real abstract for this paper. ]

[ DUMMY caption — upload your figure as CSG.png to Dr_Ajaz_and_Haji_work, or change the name here. ]

[ DUMMY BibTeX — paste the real citation here. ]

[ DUMMY — replace with the real abstract for this paper. ]

[ DUMMY caption — upload your figure as KG-EDAS-GCLR.png to Dr_Ajaz_and_Haji_work, or change the name here. ]

[ DUMMY BibTeX — paste the real citation here. ]

[ DUMMY — replace with the real abstract for this paper. ]

[ DUMMY caption — upload your figure as BERT-KGC.png to Dr_Ajaz_and_Haji_work, or change the name here. ]

[ DUMMY BibTeX — paste the real citation here. ]