Papers by Jiseong Kim

4 papers
Automatic Wordnet Mapping: from CoreNet to Princeton WordNet (L18-1)

Copied to clipboard

Challenge: Existing mappings focus on identifying the semantic categories of CoreNet, but not the word senses.
Approach: They propose to map the word senses of CoreNet into Princeton WordNet synsets by lexical relations by a taxonomy.
Outcome: The proposed mapping bridging the gap between CoreNet and WordNet shows that the word senses of CoreNet are mapped with precision of 91.2%.
Semi-automatic Korean FrameNet Annotation over KAIST Treebank (L18-1)

Copied to clipboard

Challenge: Annotating FrameNet over raw sentences is an expensive and complex task, because of which we have designed a semi-automatic annotation approach.
Approach: They propose to use Korean FrameNet annotations to build a frame-semantic parser for English using full-text annotation and partially annotated exemplar sentences to train their models.
Outcome: The proposed model is based on a lexical database of the Korean FrameNet, and its current scope, status, and limitations are discussed in the paper.
Unsupervised Korean Word Sense Disambiguation using CoreNet (L18-1)

Copied to clipboard

Challenge: Unsupervised learning based Korean word sense disambiguation is needed to distinguish between sense candidates.
Approach: They investigated unsupervised Korean word sense disambiguation using CoreNet, a Korean lexical semantic network.
Outcome: The proposed method exhibited an 80.9% accuracy on the datasets constructed and proved to be effective for practical applications.
Unsupervised Fact Checking by Counter-Weighted Positive and Negative Evidential Paths in A Knowledge Graph (2020.coling-main)

Copied to clipboard

Challenge: Misinformation spreads across media, community, and knowledge graphs in the Web by human agents and information extraction algorithms.
Approach: They propose a rule-based approach that finds positive and negative evidential paths in a knowledge graph for a given factual statement and calculates a truth score for the given statement by unsupervised ensemble.
Outcome: The proposed approach outperforms the state-of-the-art unsupervised approaches by up to 0.12 AUC-ROC and even outperfies the supervised approach by up 0.05 AUC.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations