Lab Info
Advanced Neural-nets and Topic-models (ANT) laboratory belongs to the department of Bigdata engineering in SoonChunHyang University, South Korea.
The lab has two main research area: "information extraction" and "text mining", where focused technologies are deep learning and topic modeling.
Research Area
- Korean ML, Korean TimeBank, and Temporal Information Extraction from Korean Texts
- Intelligent & Emotional Conversational System
- Out-of-domain (OOD) Detection
- CalNet: Neural Networks for Arithmetics
- Document Summarization
- Topic Comprehension by RNN
Projects
- Slot Normalizer, 2017.1 ~ 2017.6, Naver corporation
- Out-Of-Domain Detection, 2017.3 ~ 2020.2, NRF
- Enbot project, 2018.10 ~ 2019.1, Soonchunhyang Univ.
- Industry/Job code classification using deep laerning, 2018.9 ~ 2020.9, MSS
Demos
Note1: Some demo require HTTPS connection, so just accept if you want to proceed, please.
Note2: Some demo links may not be available temporarily if it is on update process.
Reports (data analysis)
- Analysis of Relationship between Length of News Article and Sentiment, 2018.3 [PDF]