The LOPEN project is a open project of Chinese language and knowledge resource maintained by LOPE (Labs of Ontologies, Language Processing and e-Humanity) at National Taiwan University. We believe open resource will enable not only reproduction and creativity of empirical research but also progress of our society.
Resource and tools
CWN Sense Tagger
In this project, we aim to solve the Chinese word sense disambiguation problem by state-of-the-art Bert model. It gives us huge performance gains and can score roughly 82% accuracy.
[link] [demo]Deep Lexicon (DeepLEX)
A large Chinese-centered open lexicon as an alternative resource to atomic lexicon theory.
[link]Chinese Wordnet (CWN)
CWN aims at constructing a deep semantic and conceptual network. Fine-grained semantic analysis and open relational design are conducive to the structure of langanguage and mine.
[link] [CWN v1] [CWN v2]HanziAnalysisKit
The Hanzi Glyph Corpus Toolkit (HGCT) and lexicoR facilitate the querying and analysis of Chinese character glyphs within corpora and provides access to various Chinese lexical resources.
[link]Chinese Word Map (CWM)
CWM is a TSCL-based (Teching Chinese as a Second Language) word sketch engine of lexical knowledge.
[link]Corpora Open and Search (COPENS)
An open corpus system and query tool. Automatically pre-processing and free annotating.
[link]PTT Corpus
As a characteristic BBS system in Taiwan, PTT records interesting social and cultural language phenomena. It also provides important empirical information on language contact and evolution.
[link]