- a better tagging method by java.
- address extraction.
- object extraction.
crf model by java.crf model by python.- a better segment implement.
cilin similarity.hownet similarity.pinyin similarity.ac+trie tree segment.double array trie tree segment.- sentiment tendency.
literal value similarity.training word2vec in java.- optimizing double arrays trie tree by storing arrays avoid building everytime.
- Aho-Corasick + double array trie to optimize matching.
- optimizing core dictionary by persistence avoid building ac-tree everytime.
- optimizing name dictionary by persistence avoid building double arrays everytime.
- optimizing idiom dictionary by persistence avoid building double arrays everytime.