Language models such as #ChatGPT can produce apparently plausible, but often incorrect scientific facts. #AI for #OpenScience requires verifiable #KnowledgeGraphs of research data. We showed their ability to produce new knowledge in our last paper: https://t.co/SwOsfj1Kuu.
#SCICERO is our last #DeepLearning and #NLP solution for extracting claims about methods, tasks, and materials from the #ScientificLiterature. Produced a #KnowledgeGraph (CS-KG, https://t.co/PDJaTtG5Q9) of 41M statements from 7M articles. Paper: https://t.co/Af6ikXbNXp
@kmiou
🏆Summer '21 to '22 most successful Springer STM books were with #ComputerScience
➡️category best-selling: The Algorithm Design Manual by Steven Skiena 👉https://t.co/FQAvdJNPoG
➡️category best-downloaded: Data Science for Economics and Finance 👉https://t.co/mH4ZHfLkxh
"You only see what you know". The starting point for a smart perception by machines using Deep Learning and Knowledge Graphs #DL4KG2019#ESWC2019@eswc_conf
At the #DL4KG2019 workshop, co-located with #ESWC2019 Mehwish Alam asks: Can KGs and DL help in Representing, Detecting and Interpreting Metaphors?
@FIZKarlsruhe@em_alam