Currently, the blacklist in mecab_wrapper.py supports reading two grammar positions, labelled pos and subPos. Allowing Mecab to read beyond that will allow filtering for more subtypes, such as:
名詞[pos],固有名詞[subPos],人名[n/a] = People's names, e.g. 田中
名詞[pos],固有名詞[subPos],組織[n/a] = Company names, e.g. 東芝
Being able to filter the above would improve the quality of cards.
Currently, the blacklist in mecab_wrapper.py supports reading two grammar positions, labelled pos and subPos. Allowing Mecab to read beyond that will allow filtering for more subtypes, such as:
Being able to filter the above would improve the quality of cards.