Boshko Koloski
@bkoloskii.bsky.social
Ph.D. student exploring the intersection of Natural Language Processing and Learning on Graphs.
🚀 Excited to share SEKE: Specialised Experts for Keyword Extraction – a novel supervised approach built on top of the Mixture-of-Experts paradigm!
📊 SOTA Performance: Excels across benchmarks and domains.
🔍 Interpretability: Reveals expert specializations.
#NLP #KeywordExtraction #MachineLearning
📊 SOTA Performance: Excels across benchmarks and domains.
🔍 Interpretability: Reveals expert specializations.
#NLP #KeywordExtraction #MachineLearning
SEKE: Specialised Experts for Keyword Extraction
Keyword extraction involves identifying the most descriptive words in a document, allowing automatic categorisation and summarisation of large quantities of diverse textual data. Relying on the insigh...
arxiv.org
December 20, 2024 at 3:36 PM
🚀 Excited to share SEKE: Specialised Experts for Keyword Extraction – a novel supervised approach built on top of the Mixture-of-Experts paradigm!
📊 SOTA Performance: Excels across benchmarks and domains.
🔍 Interpretability: Reveals expert specializations.
#NLP #KeywordExtraction #MachineLearning
📊 SOTA Performance: Excels across benchmarks and domains.
🔍 Interpretability: Reveals expert specializations.
#NLP #KeywordExtraction #MachineLearning
I love starting my Sunday mornings by reviewing well-executed minor revisions.
November 24, 2024 at 9:20 AM
I love starting my Sunday mornings by reviewing well-executed minor revisions.
trying to embed a 20-billion triplet KG results in almost ~random~ embeddings 🥴
November 17, 2024 at 6:20 PM
trying to embed a 20-billion triplet KG results in almost ~random~ embeddings 🥴