- Collection and Processing of massive Corpus
- We collect and process massive Korean-language Corpus on the web to learn understanding models and generation models of natural language.
We have news, blogs, statutes and precedents, product comments, open data and social data.
- Natural Language Understanding
- We develop deep learning technology for understanding natural language.
In particular, we actively develop natural language understanding models.
Natural language understanding models are models that allow computers to understand human speech and make it possible for document classification, conversational understanding, machine reading comprehension and emotional analysis.
We build vocabulary dictionaries and trained models specialized for Korean language.
- Natural Language Generation
- Based on deep learning, we have developed the summarizer which sums up various documents such as news and blogs.