Jan Portisch
@janportisch
Development Architect @SAP
Ph.D. in CS @dwsunima
ID: 1142380009589497856
https://www.jan-portisch.eu/ 22-06-2019 10:33:13
71 Tweet
120 Followers
207 Following
The zoo of #rdf2vec variants is constantly growing. Our ESWC Conferences poster, to be presented at #eswc2022, provides a preliminary comparison of 12 variants, using different walk strategies and #word2vec training modalities. arxiv.org/abs/2204.02777 w/ Jan Portisch Data and Web Science Group
Our survey on #BackgroundKnowledge in #OntologyMatching is accepted at Semantic Web Journal w/ Jan Portisch Michael Hladik Data and Web Science Group :-) Overall, we reviewed more than 300 papers discussing background knowledge in ontology matching. Key findings of the survey in this thread: (1/7)
Our poster on #ontologymatching through absolute orientation of #embedding spaces has been accepted at ESWC Conferences! Preprint here: arxiv.org/abs/2204.04040 Key take away: structurally similar graphs can be aligned by rotation of their corresponding latent embedding spaces.
#Transformer models do cool things in NLP, and they can also help reducing the search space in #ontology, #schema, and #knowledgegraph #matching. 👉"KERMIT 🐸 A Transformer-Based Approach for Knowledge Graph Matching" arxiv.org/abs/2204.13931 Sven Hertling Jan Portisch Data and Web Science Group
#eswc2022 was quite a success for Data and Web Science Group: best poster award for Jan Portisch and me, and both @n_heist and myself being recognized as outstanding reviewers :-) 2022.eswc-conferences.org/awards/
Happy that our work with Russa Biswas on "Entity Type Prediction Leveraging Graph Walks and Entity Descriptions" has been accepted International Semantic Web Conference #iswc2022 w/ Jan Portisch Heiko Paulheim Harald Sack Preprint of the paper will soon be available. #machinelearning #knowledgegraphs
Our #iswc2022 research track paper on "Entity Type Prediction Leveraging Graph Walks and Entity Descriptions" by Russa Biswas Jan Portisch Heiko Paulheim Harald Sack Mehwish Alam is now available on arxiv.org/abs/2207.14094 #knowledgeGraphs International Semantic Web Conference FIZ Karlsruhe
Congratulations to Jan Portisch, who sucessfully defended his thesis on "Exploiting General-Purpose #BackgroundKnowledge for Automated #SchemaMatching" yesterday at Data and Web Science Group in front of the jury: Catia Pesquita @hanvdaa and Christian Bizer!
"Gollum: A Gold Standard for Large Scale Multi Source Knowledge Graph #Matching" is accepted for AKBC 2022, introducing a gold standard of 275k correspondences between 4k #knowledgegraphs. Paper: arxiv.org/abs/2209.07479 Data: doi.org/10.5281/zenodo… Sven Hertling Data and Web Science Group
I will present our work on the DLCC benchmark for knowledge graph embeddings in the embeddings 2 session at International Semantic Web Conference #iswc2022 in 45 minutes. Paper: arxiv.org/abs/2207.06014 Code: github.com/janothan/DL-TC… Data: zenodo.org/record/6509715 w/ Jan Portisch Data and Web Science Group
Our colleague Russa Biswas will present our paper "Entity Type Prediction Leveraging Graph Walks and Entity Descriptions" today at International Semantic Web Conference #iswc2022 joint work with Heiko Paulheim Jan Portisch Mehwish Alam Harald Sack FIZ Karlsruhe fiz-karlsruhe.de/sites/default/…
Finally, #RDF2vec is becoming a N̵e̵t̵f̵l̵i̵x̵ ̵s̵e̵r̵i̵e̵s̵ book at Springer Nature, co-authored by Jan Portisch, Petar Ristoski, and me. link.springer.com/book/10.1007/9…
Very proud that Data and Web Science Group alumnus Jan Portisch receives the SWSA distinguished dissertation award. Well done, Jan!
Our #rdf2vec book is available at the Springer Nature booth at International Semantic Web Conference #iswc2023 Jan Portisch Petar Ristoski
#LLMs are excellent few-shot #ontology #matching systems. Read our #kcap2023 paper here: arxiv.org/abs/2311.03837 w/ Sven Hertling Data and Web Science Group