Kronecker Decomposition for Knowledge Graph Embeddings

Demir, Caglar; Lienen, Julian; Ngomo, Axel-Cyrille Ngonga

doi:10.1145/3511095.3531276

Cited by 4 publications

(6 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We used the most recent DBpedia 2021 benchmark dataset 3 to evaluate our framework in depth. Our experiments suggest that a state-of-the-art KGE model with more than 11.4B parameters can be successfully trained and applied in link prediction, and relation prediction 4 . We refer to the project page for the details and log files about pretrained models.…”

Section: Summary Of Initial Experimental Resultsmentioning

confidence: 94%

“…Scholarly Publications: Our framework have been effectively used to learn knowledge graphs embeddings in several published works [13], [3], [14], [4], [15], [16], and [17].…”

Section: Software Impactmentioning

confidence: 99%

“…Given the triples (Barack, Married, Michelle) and (Michelle, HasChild, Malia) ∈ G, a good scoring function is expected to return high scores for (Barack, HasChild, Malia) and (Michelle, Married, Barack), while returning a considerably lower score for (Malia, HasChild, Barack). To compute a single score, embeddings of entities and relations are retrieved from E, R and trilinear d-dimension vector multiplication is performed, i.e., φ(Barack, HasChild, Malia) = Barack • HasChild • Malia (see [4]).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Hardware-agnostic Computation for Large-scale Knowledge Graph Embeddings

Demir¹,

Ngomo²

2022

Preprint

Self Cite

View full text Add to dashboard Cite

Knowledge graph embedding research has mainly focused on learning continuous representations of knowledge graphs towards the link prediction problem. Recently developed frameworks can be effectively applied in research related applications. Yet, these frameworks do not fulfill many requirements of real-world applications. As the size of the knowledge graph grows, moving computation from a commodity computer to a cluster of computers in these frameworks becomes more challenging. Finding suitable hyperparameter settings w.r.t. time and computational budgets are left to practitioners. In addition, the continual learning aspect in knowledge graph embedding frameworks is often ignored, although continual learning plays an important role in many real-world (deep) learning-driven applications. Arguably, these limitations explain the lack of publicly available knowledge graph embedding models for large knowledge graphs. We developed a framework based on the frameworks DASK, Pytorch Lightning and Hugging Face to compute embeddings for large-scale knowledge graphs in a hardware-agnostic manner, which is able to address real-world challenges pertaining to the scale of real application. We provide an open-source version of our framework along with a hub of pre-trained models having more than 11.4 B parameters 2 .

show abstract

Section: Summary Of Initial Experimental Resultsmentioning

confidence: 94%

“…Scholarly Publications: Our framework have been effectively used to learn knowledge graphs embeddings in several published works [13], [3], [14], [4], [15], [16], and [17].…”

Section: Software Impactmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Hardware-agnostic Computation for Large-scale Knowledge Graph Embeddings

Demir¹,

Ngomo²

2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Scholarly Publications: Our framework have been effectively used to learn knowledge graphs embeddings in several published works [3,4,[13][14][15][16], and [17].…”

Section: Software Impactmentioning

confidence: 99%

“…Given the triples (Barack, Married, Michelle) and (𝙼𝚒𝚌𝚑𝚎𝚕𝚕𝚎, 𝙷𝚊𝚜𝙲𝚑𝚒𝚕𝚍, 𝙼𝚊𝚕𝚒𝚊) ∈ , a good scoring function is expected to return high scores for (Barack, HasChild, Malia) and (Michelle, Married, Barack), while returning a considerably lower score for (Malia, HasChild, Barack). To compute a single score, embeddings of entities and relations are retrieved from 𝐄, 𝐑 and trilinear d-dimension vector multiplication is performed, i.e., 𝜙(𝙱𝚊𝚛𝚊𝚌𝚔, 𝙷𝚊𝚜𝙲𝚑𝚒𝚕𝚍, 𝙼𝚊𝚕𝚒𝚊) = 𝐁𝐚𝐫𝐚𝐜𝐤 • 𝐇𝐚𝐬𝐂𝐡𝐢𝐥𝐝 ⋅ 𝐌𝐚𝐥𝐢𝐚 (see [4]).…”

Section: Introductionmentioning

confidence: 99%