S
Sunghwan Mac Kim
Researcher at Commonwealth Scientific and Industrial Research Organisation
Publications - 14
Citations - 459
Sunghwan Mac Kim is an academic researcher from Commonwealth Scientific and Industrial Research Organisation. The author has contributed to research in topics: Probabilistic latent semantic analysis & Latent semantic analysis. The author has an hindex of 7, co-authored 14 publications receiving 400 citations. Previous affiliations of Sunghwan Mac Kim include University of Sydney.
Papers
More filters
Journal ArticleDOI
Emotions in text: dimensional and categorical models
Rafael A. Calvo,Sunghwan Mac Kim +1 more
TL;DR: A new way of using normative databases as a way of processing text with a dimensional model and compare it with different categorical approaches is introduced and shows that the categorical model using NMF and the dimensional model tend to perform best.
Proceedings Article
Evaluation of Unsupervised Emotion Models to Textual Affect Recognition
TL;DR: Experiments show that a categorical model using NMF results in better performances for SemEval and fairy tales, whereas a dimensional model performs better with ISEAR.
Proceedings ArticleDOI
Demographic Inference on Twitter using Recursive Neural Networks
TL;DR: This work employs recursive neural networks to break down independence assumptions to obtain inference about demographic characteristics on Twitter and shows that this model performs better than existing models including the state-of-the-art.
Proceedings Article
Sentiment Analysis in Student Experiences of Learning.
Sunghwan Mac Kim,Rafael A. Calvo +1 more
TL;DR: An evaluation of new techniques for automatically detecting sentiment polarity (Positive or Negative) in the students responses to Unit of Study Evaluations (USE) finds NMF-based categorical model and dimensional model result in better performances above the baseline.
Finding Names in Trove: Named Entity Recognition for Australian Historical Newspapers
Sunghwan Mac Kim,Steve Cassidy +1 more
TL;DR: An evaluation of the Stanford NER system on this data is presented and some analysis of the results including a version published as Linked Data and an exploration of clustering the mentions of certain names in the archive to try to identify individuals.